Podcast appearances and mentions of Andrew Lo

Javier Ruiz es gestor en Horos Asset Management, un fondo de inversión que sigue la metodología del value investing, que podría resumirse como comprar buenos productos a buenos precios. Entendí su filosofía cuando me contó la tesis del uranio. Los mercados presentan ineficiencias y los gestores activos intentan aprovecharse de ellas. Las oportunidades de arbitraje son incluso mayores con la popularidad reciente de la inversión pasiva. Javier detalla los aprendizajes en sus magníficas cartas trimestrales.Kapital es posible gracias a sus colaboradores:⁠⁠⁠La casa ESE⁠⁠⁠. ¿Cómo quieres vivir?Aquí de vuelta los pesaos queridos amigos de La casa ESE. Buscando la forma de seguir inventando cosas ya inventadas hemos creado mapadecasas.com, allí tendréis la oportunidad de encontrar, más que vuestra futura casa, vuestra futura vida. Sí, es muy ambicioso. En Madrid, por ejemplo, vamos a crear un conjunto residencial donde además de habitar, podamos llevar un poquito del Mediterráneo moral. No sólo una casa, sino un lugar que tenga zonas verdes, espacios comunitarios y hasta un edificio que pueda hacer las veces de coworking entre otras cosas. A 30 minutos de Madrid y buscando gente afín al mundo tecnológico, al emprendimiento, al marketing y a la cultura. Visita la propuesta de Distrito ESE.⁠UTAMED⁠. La universidad online del siglo XXI.UTAMED, la universidad oficial y online de la Fundación Unicaja, nace para romper las barreras que durante décadas han limitado el acceso a la educación y la cultura. Con exámenes 100 % online y financiación sin intereses, ofrecemos una formación accesible, flexible y comprometida con el presente. Porque hoy ya no basta con obtener un título: en UTAMED te preparamos para trabajar desde el primer año. Lo hacemos junto a la empresa, adaptando los contenidos académicos a sus demandas reales, para que nuestros estudiantes adquieran las competencias más valoradas en el mercado laboral. Por ser oyente de este podcast, tienes un descuento del 30% en todo el catálogo de grados y másteres, oficiales y propios.Patrocina Kapital. Toda la información en este link.Índice:2:25 La estafa del Forúm Filatélico.9:28 Rivalidades personales en los mercados.17:43 Munger clasifica 25 errores de comportamiento.20:11 La última decisión de Kahneman.30:01 El instinto maladaptado de seguir al rebaño.35:21 Apalancamiento mortal.40:54 Fuentes de valor añadido según Mauboussin.53:26 Invertir en compañías aburridas.1:01:58 Comisiones en la gestión de activa.1:12:45 Explicárselo a un niño de cinco años.1:15:56 Refritos de ETFs con comisión del 2%.1:26:50 Objetivar el proceso de decisión.1:32:49 La paradoja del margen de seguridad.1:36:19 El criterio de Kelly.1:41:25 La fantástica tesis del uranio.1:51:37 Ampliación de capital para pagar dividendo.1:57:34 Anticipar el sentimiento colectivo.2:01:57 La historia de los tipos de interés.2:06:17 Teoría austríaca del ciclo económico.Apuntes:Measuring the moat. Michael Mauboussin & Dan Callahan.The adaptive market hypothesis. Andrew Lo.El enigma de la experiencia frente a la memoria. Daniel Kahneman.Cartas a los accionistas. Seth Klarman.Herbalife. Bill Ackman.Un paso por delante de Wall Street. Peter Lynch.The model. Richard Lawrence.El diccionario financiero del diablo. Jason Zweig.A man for all markets. Edward Thorp.Rendimientos del capital. Edward Chancellor.El precio del tiempo. Edward Chancellor.

The Registered Education Savings Plan is a home run when it comes to savin g for your kids education. There are various ways to set up an RESP and we have Andrew Lo, CEO of Embark, joins us to talk about how they go about it. Connect with Embark on Facebook, LinkedIn, Instagram, TikTok and YouTube.

Retiring abroad. Jen Barnett, co-founder of Expatsi, takes us through the financial considerations people have to make before spending retirement in a new country. Then, saving for your child's education during times of economic uncertainty. Andrew Lo, CEO of Embark, helps us navigate those financial challenges. And, going through a divorce when you have joint debt. Wesley Cowan, senior vice-president and licensed insolvency trustee with MNP, tells us what happens to this debt during insolvency. Plus, we speak with Drew Boyer, certified financial planner and hip-hop enthusiast, about his book HIP HOP X FINANCE: Become a Financial Gangster, Get Off Debt Row and Stack Your Cash Flow. To find out more about the guests check out: Andrew Lo: Facebook | LinkedIn | Instagram | TikTok | YouTube Drew Boyer: hiphopxfinance.com | Facebook | TikTok | Instagram Wesley Cowan: MNP | LinkedIn MNP: mnpdebt.ca Jen Barnett: Facebook | Instagram | TikTok Bruce Sellery is a personal finance expert and best-selling author. As the founder of Moolala and the CEO of Credit Canada, Bruce is on a mission to help you get a better handle on your money so you can live the life you want. High energy & low B.S., this is Moolala: Money Made Simple. Find Bruce Sellery at Moolala.ca | Twitter | Facebook | LinkedIn

Plastic pollution clean up through absorption is now a possibility by using a combination of cotton and chittin to take microplastics out of the ocean...but can this work at scale? In this episode of the "How to Protect the Ocean" podcast, host Andrew Lo discusses the pressing issue of microplastics in the ocean and introduces a promising new method for their absorption using a combination of cotton and chitin. The mechanism involves creating a sponge-like substance called CT cell biomass, which combines cellulose from cotton and chitin from squid. This innovative material is designed to effectively capture microplastics from water. The process begins by breaking the original hydrogen bonds in cellulose and chitin, allowing them to bind together and form a stable framework with numerous activated hydrogen bonding sites. This structure enhances the material's ability to absorb microplastics through various interactions, including physical interception and electrostatic attraction. Research indicates that this foam can remove 98 to 99.9% of microplastics from water samples, showcasing its potential as an eco-friendly solution for addressing microplastic pollution. The episode emphasizes the importance of developing sustainable strategies for microplastic remediation in aquatic environments, while also highlighting the need for broader efforts to reduce plastic usage at the source. Link to article: https://www.iflscience.com/new-sponge-like-biomass-foam-found-to-soak-up-999-percent-of-microplastics-77223 Follow a career in conservation: https://www.conservation-careers.com/online-training/ Use the code SUFB to get 33% off courses and the careers program.   Do you want to join my Ocean Community? Sign Up for Updates on the process: www.speakupforblue.com/oceanapp   Sign up for our Newsletter: http://www.speakupforblue.com/newsletter   Facebook Group: https://bit.ly/3NmYvsI Connect with Speak Up For Blue: Website: https://bit.ly/3fOF3Wf Instagram: https://bit.ly/3rIaJSG TikTok: https://www.tiktok.com/@speakupforblue Twitter: https://bit.ly/3rHZxpc YouTube: www.speakupforblue.com/youtube  

Thank you Brytn Beggs , Andrew Lo, Kia Mazhar and Alex Awad. Shoutout the roommate Mo!

Professor of Finance at the MIT Sloan School of Management and CSAIL Andrew Lo believes AI can help everyday consumers make important financial decisions by democratizing access to quality finance advice. His research aims to address the challenges of deploying AI in finance by, for example, answering questions around responsibility and engaging with financial advisors to make sure such tools are useful in the field. Professor Lo is the faculty director for the FintechAI@CSAIL research initiative. Find out more about CSAIL Alliances, as well as a full transcript of this podcast, at https://cap.csail.mit.edu/podcasts/how-ai-can-help-financial-decision-making-andrew-lo If you would like to learn more about CSAIL's Professional Development Courses, including the upcoming Driving Innovation with Generative AI, visit here: cap.csail.mit.edu/events-professional-programs. Podcast listeners save 10% on courses with code MITXPOD10. Looking for another great podcast? MIT Sloan Management Review's "Me, Myself, and AI," expert hosts and researchers talk with AI leaders from organizations like NASA, Upwork, Github, and Meta to explore how organizations achieve success with generative AI — and what challenges and ethical considerations they face along the way. Listen to Me, Myself, and AI wherever you stream podcasts. https://link.chtbl.com/pxsEZ4pf?sid=CSAIL

IN THIS EPISODE: In this podcast episode, host Philip Guarino speaks with Andrew Lo, a distinguished professor at MIT and an entrepreneur with a focus on healthcare and deep tech ventures. Andrew shares his journey from academia to entrepreneurship, detailing his transition from economics to founding a quantitative investment firm and later exploring the financial challenges of drug development. He discusses the fundraising challenges for deep tech entrepreneurs, how investors' risk calculation differs for more complex ventures and how this requires continuous investor communication and learning. Lo also highlights the vibrant innovation ecosystem in the Boston-Cambridge area and his passion for mentoring students in entrepreneurial projects. GUEST BIO: Andrew W. Lo is the Charles E. and Susan T. Harris Professor at the MIT Sloan School of Management, director of MIT's Laboratory for Financial Engineering, and principal investigator at MIT's Computer Science and Artificial Intelligence Laboratory. He is a co-founder and director of BridgeBio Pharma, a director of AbCellera, Atomwise, and Vesalius, a co-founder and chairman of QLS Advisors, and a member of the advisory board to the American Cancer Society's BrightEdge Impact Fund. Dr. Lo received his B.A. in economics from Yale University and his A.M. and Ph.D. in economics from Harvard University.

The war between Israel and terrorist group Hamas has dominated headlines for over 6 months, but with Iran now taking aim at Israel, is there a threat that this conflict could spreading into the neighboring region? We get the latest on the conflict including the potential for the use of nuclear weapons with Mercedes Stephenson, Global News Ottawa Bureau Chief and Host of “The West Block”. Next, the price of post-secondary education is going up year after year. According to Stats Canada, the average yearly cost for Canadian undergraduates has increased to just over $7,000 dollars, up 3% from the 2022-2023 academic year. Are you prepared to help you kids navigate the cost of a post-secondary education, and are you willing take on debt to do it? We tackle the topic with Andrew Lo, CEO and President of “Embark” Canada.   Finally, the Coachella music festival wrapped up its second weekend with attendance hitting around 300 000 people over the two weekends, the highest attendance the festival has ever seen according to sources. But is this an outlier? Has enthusiasm for music festivals started to die away? We get the thoughts of Alan Cross, Host of “The Ongoing History of New Music” on the Corus Radio Network.

Martin talks to Andrew Lo, CEO of Embark, on the importance of RESPs and how to save for your child's education.

Andrew Lo, Professor of Finance, and the Director of the Laboratory for Financial Engineering at the MIT Sloan School of Management, sits down with Neil to discuss his application of portfolio theory to address the challenges of funding drug development and how a Netflix subscription model can address the coming crisis in funding high-priced curative therapies approaching the market.

Hi everyone. We're taking the week off for the 4th of July holiday, but we wanted to use this week's episode to honor Nobel Prize-winning economist Harry Markowitz, who recently passed away at the age of 95. Professor Markowitz is a giant of finance, someone who put diversification and Modern Portfolio Theory on the map, with his research transforming the way we allocate and invest our assets. While we didn't have the opportunity to interview Professor Markowitz for the podcast, we were able to chat recently with someone who had interviewed him: author and financial researcher Dr. Andrew Lo. Dr. Lo recently published a book titled “In Pursuit of the Perfect Portfolio,” in which he profiled some of the leading figures in academic research and finance. None stood taller than Professor Markowitz, whom Dr. Lo discusses at length in this interview we aired in February of 2022. We think you'll enjoy it. Thanks so much for listening and see you in a week. Have a happy holiday.Our guest this week is Dr. Andrew Lo. Dr. Lo is the Charles E. & Susan T. Harris Professor, a professor of finance, and the director of the Laboratory for Financial Engineering at the MIT Sloan School of Management. His current research spans five areas, including evolutionary models of investor behavior and adaptive markets, systemic risk, and financial regulation, among others. Dr. Lo has published extensively in academic journals and authored a number of books including In Pursuit of the Perfect Portfolio, which he cowrote with Stephen Foerster. He has received numerous awards for his work and contributions to modern finance research throughout his career. He holds a bachelor's in economics from Yale University and an AM and Ph.D. in economics from Harvard University.BackgroundIn Pursuit of the Perfect Portfolio: The Stories, Voices, and Key Insights of the Pioneers Who Shaped the Way We Invest, by Andrew W. Lo and Stephen R. FoersterAdaptive Markets: Financial Evolution at the Speed of Thought, by Andrew W. LoHistory"Thirty Maidens of Geneva," the Tontine Coffee-House, thetch.blog.com, Aug. 5, 2019."Why 18th Century Swiss Bankers Bet on the Lives of Young Girls," by Stephen Foerster, sfoerster-5338.medium.com, Sept. 2, 2021.William F. Sharpe"Keynes the Stock Market Investor: A Quantitative Analysis," by David Chambers, Elroy Dimson, and Justin Foo, papers.ssrn.com, Sept. 26, 2013.Eugene F. Fama"Algorithmic Models of Investor Behavior," by Andrew Lo and Alexander Remorov, eqderivatives.com, 2021."In Pursuit of the Perfect Portfolio: Eugene Fama," Interview with Andrew Lo and Eugene Fama, youtube.com, Dec. 15, 2016."Why Artificial Intelligence May Not Be as Useful or as Challenging as Artificial Stupidity," by Andrew Lo, hdsr.mitpress.mit.edu, July 1, 2019.Charles D. Ellis"Charley Ellis: Why Active Investing Is Still a Loser's Game," The Long View podcast, Morningstar.com, May 27, 2020.Other"7 Principles to Help You Create Your Perfect Portfolio," by Robert Powell, marketwatch.com, Nov. 10, 2021. 

Andrew Lo, president and CEO of Embark Learn more about your ad choices. Visit megaphone.fm/adchoices

Latent Space is popping off! Welcome to the over 8500 latent space explorers who have joined us. Join us this month at various events in SF and NYC, or start your own!This post spent 22 hours at the top of Hacker News.As announced during their Developer Day celebrating their $100m fundraise following their Google partnership, Replit is now open sourcing its own state of the art code LLM: replit-code-v1-3b (model card, HF Space), which beats OpenAI's Codex model on the industry standard HumanEval benchmark when finetuned on Replit data (despite being 77% smaller) and more importantly passes AmjadEval (we'll explain!)We got an exclusive interview with Reza Shabani, Replit's Head of AI, to tell the story of Replit's journey into building a data platform, building GhostWriter, and now training their own LLM, for 22 million developers!8 minutes of this discussion go into a live demo discussing generated code samples - which is always awkward on audio. So we've again gone multimodal and put up a screen recording here where you can follow along on the code samples!Recorded in-person at the beautiful StudioPod studios in San Francisco.Full transcript is below the fold. We would really appreciate if you shared our pod with friends on Twitter, LinkedIn, Mastodon, Bluesky, or your social media poison of choice!Timestamps* [00:00:21] Introducing Reza* [00:01:49] Quantitative Finance and Data Engineering* [00:11:23] From Data to AI at Replit* [00:17:26] Replit GhostWriter* [00:20:31] Benchmarking Code LLMs* [00:23:06] AmjadEval live demo* [00:31:21] Aligning Models on Vibes* [00:33:04] Beyond Chat & Code Completion* [00:35:50] Ghostwriter Autonomous Agent* [00:38:47] Releasing Replit-code-v1-3b* [00:43:38] The YOLO training run* [00:49:49] Scaling Laws: from Kaplan to Chinchilla to LLaMA* [00:52:43] MosaicML* [00:55:36] Replit's Plans for the Future (and Hiring!)* [00:59:05] Lightning RoundShow Notes* Reza Shabani on Twitter and LinkedIn* also Michele Catasta and Madhav Singhal* Michele Catasta's thread on the release of replit-code-v1-3b* Intro to Replit Ghostwriter* Replit Ghostwriter Chat and Building Ghostwriter Chat* Reza on how to train your own LLMs (their top blog of all time)* Our Benchmarks 101 episode where we discussed HumanEval* AmjadEval live demo* Nat.dev* MosaicML CEO Naveen Rao on Replit's LLM* MosaicML Composer + FSDP code* Replit's AI team is hiring in North America timezone - Fullstack engineer, Applied AI/ML, and other roles!Transcript[00:00:00] Alessio Fanelli: Hey everyone. Welcome to the Latent Space podcast. This is Alessio, partner and CTO in residence at Decibel Partners. I'm joined by my co-host, swyx, writer and editor of Latent Space.[00:00:21] Introducing Reza[00:00:21] swyx: Hey and today we have Reza Shabani, Head of AI at Replit. Welcome to the studio. Thank you. Thank you for having me. So we try to introduce people's bios so you don't have to repeat yourself, but then also get a personal side of you.[00:00:34] You got your PhD in econ from Berkeley, and then you were a startup founder for a bit, and, and then you went into systematic equity trading at BlackRock in Wellington. And then something happened and you were now head of AI at Relet. What should people know about you that might not be apparent on LinkedIn?[00:00:50] One thing[00:00:51] Reza Shabani: that comes up pretty often is whether I know how to code. Yeah, you'd be shocked. A lot of people are kind of like, do you know how to code? When I was talking to Amjad about this role, I'd originally talked to him, I think about a product role and, and didn't get it. Then he was like, well, I know you've done a bunch of data and analytics stuff.[00:01:07] We need someone to work on that. And I was like, sure, I'll, I'll do it. And he was like, okay, but you might have to know how to code. And I was like, yeah, yeah, I, I know how to code. So I think that just kind of surprises people coming from like Ancon background. Yeah. Of people are always kind of like, wait, even when people join Relet, they're like, wait, does this guy actually know how to code?[00:01:28] Is he actually technical? Yeah.[00:01:30] swyx: You did a bunch of number crunching at top financial companies and it still wasn't[00:01:34] Reza Shabani: obvious. Yeah. Yeah. I mean, I, I think someone like in a software engineering background, cuz you think of finance and you think of like calling people to get the deal done and that type of thing.[00:01:43] No, it's, it's not that as, as you know, it's very very quantitative. Especially what I did in, in finance, very quantitative.[00:01:49] Quantitative Finance and Data Engineering[00:01:49] swyx: Yeah, so we can cover a little bit of that and then go into the rapid journey. So as, as you, as you know, I was also a quantitative trader on the sell side and the buy side. And yeah, I actually learned Python there.[00:02:01] I learned my, I wrote my own data pipelines there before airflow was a thing, and it was just me writing running notebooks and not version controlling them. And it was a complete mess, but we were managing a billion dollars on, on my crappy code. Yeah, yeah. What was it like for you?[00:02:17] Reza Shabani: I guess somewhat similar.[00:02:18] I, I started the journey during grad school, so during my PhD and my PhD was in economics and it was always on the more data intensive kind of applied economic side. And, and specifically financial economics. And so what I did for my dissertation I recorded cnbc, the Financial News Network for 10 hours a day, every day.[00:02:39] Extracted the close captions from the video files and then used that to create a second by second transcript of, of cmbc, merged that on with high frequency trading, quote data and then looked at, you know, went in and did some, some nlp, tagging the company names, and and then looked at the price response or the change in price and trading volume in the seconds after a company was mentioned.[00:03:01] And, and this was back in. 2009 that I was doing this. So before cloud, before, before a lot of Python actually. And, and definitely before any of these packages were available to make this stuff easy. And that's where, where I had to really learn to code, like outside of you know, any kind of like data programming languages.[00:03:21] That's when I had to learn Python and had to learn all, all of these other skills to work it with data at that, at that scale. So then, you know, I thought I wanted to do academia. I did terrible on the academic market because everyone looked at my dissertation. They're like, this is cool, but this isn't economics.[00:03:37] And everyone in the computer science department was actually way more interested in it. Like I, I hung out there more than in the econ department and You know, didn't get a single academic offer. Had two offer. I think I only applied to like two industry jobs and got offers from both of them.[00:03:53] They, they saw value in it. One of them was BlackRock and turned it down to, to do my own startup, and then went crawling back two and a half years later after the startup failed.[00:04:02] swyx: Something on your LinkedIn was like you're trading Chinese news tickers or something. Oh, yeah. I forget,[00:04:07] Reza Shabani: forget what that was.[00:04:08] Yeah, I mean oh. There, there was so much stuff. Honestly, like, so systematic active equity at, at BlackRock is, was such an amazing. Group and you just end up learning so much and the, and the possibilities there. Like when you, when you go in and you learn the types of things that they've been trading on for years you know, like a paper will come out in academia and they're like, did you know you can use like this data on searches to predict the price of cars?[00:04:33] And it's like, you go in and they've been trading on that for like eight years. Yeah. So they're, they're really ahead of the curve on, on all of that stuff. And the really interesting stuff that I, that I found when I went in was all like, related to NLP and ml a lot of like transcript data, a lot of like parsing through the types of things that companies talk about, whether an analyst reports, conference calls, earnings reports and the devil's really in the details about like how you make sense of, of that information in a way that, you know, gives you insight into what the company's doing and, and where the market is, is going.[00:05:08] I don't know if we can like nerd out on specific strategies. Yes. Let's go, let's go. What, so one of my favorite strategies that, because it never, I don't think we ended up trading on it, so I can probably talk about it. And it, it just kind of shows like the kind of work that you do around this data.[00:05:23] It was called emerging technologies. And so the whole idea is that there's always a new set of emerging technologies coming onto the market and the companies that are ahead of that curve and stay up to date on on the latest trends are gonna outperform their, their competitors.[00:05:38] And that's gonna reflect in the, in the stock price. So when you have a theory like that, how do you actually turn that into a trading strategy? So what we ended up doing is, well first you have to, to determine what are the emergent technologies, like what are the new up and coming technologies.[00:05:56] And so we actually went and pulled data on startups. And so there's like startups in Silicon Valley. You have all these descriptions of what they do, and you get that, that corpus of like when startups were getting funding. And then you can run non-negative matrix factorization on it and create these clusters of like what the various Emerging technologies are, and you have this all the way going back and you have like social media back in like 2008 when Facebook was, was blowing up.[00:06:21] And and you have things like mobile and digital advertising and and a lot of things actually outside of Silicon Valley. They, you know, like shale and oil cracking. Yeah. Like new technologies in, in all these different types of industries. And then and then you go and you look like, which publicly traded companies are actually talking about these things and and have exposure to these things.[00:06:42] And those are the companies that end up staying ahead of, of their competitors. And a lot of the the cases that came out of that made a ton of sense. Like when mobile was emerging, you had Walmart Labs. Walmart was really far ahead in terms of thinking about mobile and the impact of mobile.[00:06:59] And, and their, you know, Sears wasn't, and Walmart did well, and, and Sears didn't. So lots of different examples of of that, of like a company that talks about a new emerging trend. I can only imagine, like right now, all of the stuff with, with ai, there must be tons of companies talking about, yeah, how does this affect their[00:07:17] swyx: business?[00:07:18] And at some point you do, you do lose the signal. Because you get overwhelmed with noise by people slapping a on everything. Right? Which is, yeah. Yeah. That's what the Long Island Iced Tea Company slaps like blockchain on their name and, you know, their stock price like doubled or something.[00:07:32] Reza Shabani: Yeah, no, that, that's absolutely right.[00:07:35] And, and right now that's definitely the kind of strategy that would not be performing well right now because everyone would be talking about ai. And, and that's, as you know, like that's a lot of what you do in Quant is you, you try to weed out other possible explanations for for why this trend might be happening.[00:07:52] And in that particular case, I think we found that, like the companies, it wasn't, it wasn't like Sears and Walmart were both talking about mobile. It's that Walmart went out of their way to talk about mobile as like a future, mm-hmm. Trend. Whereas Sears just wouldn't bring it up. And then by the time an invest investors are asking you about it, you're probably late to the game.[00:08:12] So it was really identifying those companies that were. At the cutting edge of, of new technologies and, and staying ahead. I remember like Domino's was another big one. Like, I don't know, you[00:08:21] swyx: remember that? So for those who don't know, Domino's Pizza, I think for the run of most of the 2010s was a better performing stock than Amazon.[00:08:29] Yeah.[00:08:31] Reza Shabani: It's insane.[00:08:32] swyx: Yeah. Because of their investment in mobile. Mm-hmm. And, and just online commerce and, and all that. I it must have been fun picking that up. Yeah, that's[00:08:40] Reza Shabani: that's interesting. And I, and I think they had, I don't know if you, if you remember, they had like the pizza tracker, which was on, on mobile.[00:08:46] I use it[00:08:46] swyx: myself. It's a great, it's great app. Great app. I it's mostly faked. I think that[00:08:50] Reza Shabani: that's what I heard. I think it's gonna be like a, a huge I don't know. I'm waiting for like the New York Times article to drop that shows that the whole thing was fake. We all thought our pizzas were at those stages, but they weren't.[00:09:01] swyx: The, the challenge for me, so that so there's a, there's a great piece by Eric Falkenstein called Batesian Mimicry, where every signal essentially gets overwhelmed by noise because the people who wants, who create noise want to follow the, the signal makers. So that actually is why I left quant trading because there's just too much regime changing and like things that would access very well would test poorly out a sample.[00:09:25] And I'm sure you've like, had a little bit of that. And then there's what was the core uncertainty of like, okay, I have identified a factor that performs really well, but that's one factor out of. 500 other factors that could be going on. You have no idea. So anyway, that, that was my existential uncertainty plus the fact that it was a very highly stressful job.[00:09:43] Reza Shabani: Yeah. This is a bit of a tangent, but I, I think about this all the time and I used to have a, a great answer before chat came out, but do you think that AI will win at Quant ever?[00:09:54] swyx: I mean, what is Rentech doing? Whatever they're doing is working apparently. Yeah. But for, for most mortals, I. Like just waving your wand and saying AI doesn't make sense when your sample size is actually fairly low.[00:10:08] Yeah. Like we have maybe 40 years of financial history, if you're lucky. Mm-hmm. Times what, 4,000 listed equities. It's actually not a lot. Yeah, no, it's,[00:10:17] Reza Shabani: it's not a lot at all. And, and constantly changing market conditions and made laden variables and, and all of, all of that as well. Yeah. And then[00:10:24] swyx: retroactively you're like, oh, okay.[00:10:26] Someone will discover a giant factor that, that like explains retroactively everything that you've been doing that you thought was alpha, that you're like, Nope, actually you're just exposed to another factor that you're just, you just didn't think about everything was momentum in.[00:10:37] Yeah. And one piece that I really liked was Andrew Lo. I think he had from mit, I think he had a paper on bid as Spreads. And I think if you, if you just. Taken, took into account liquidity of markets that would account for a lot of active trading strategies, alpha. And that was systematically declined as interest rates declined.[00:10:56] And I mean, it was, it was just like after I looked at that, I was like, okay, I'm never gonna get this right.[00:11:01] Reza Shabani: Yeah. It's a, it's a crazy field and I you know, I, I always thought of like the, the adversarial aspect of it as being the, the part that AI would always have a pretty difficult time tackling.[00:11:13] Yeah. Just because, you know, there's, there's someone on the other end trying to out, out game you and, and AI can, can fail in a lot of those situations. Yeah.[00:11:23] swyx: Cool.[00:11:23] From Data to AI at Replit[00:11:23] Alessio Fanelli: Awesome. And now you've been a rep almost two years. What do you do there? Like what does the, the team do? Like, how has that evolved since you joined?[00:11:32] Especially since large language models are now top of mind, but, you know, two years ago it wasn't quite as mainstream. So how, how has that evolved?[00:11:40] Reza Shabani: Yeah, I, so when I joined, I joined a year and a half ago. We actually had to build out a lot of, of data pipelines.[00:11:45] And so I started doing a lot of data work. And we didn't have you know, there, there were like databases for production systems and, and whatnot, but we just didn't have the the infrastructure to query data at scale and to process that, that data at scale and replica has tons of users tons of data, just tons of ripples.[00:12:04] And I can get into, into some of those numbers, but like, if you wanted to answer the question, for example of what is the most. Forked rep, rep on rep, you couldn't answer that back then because it, the query would just completely time out. And so a lot of the work originally just went into building data infrastructure, like modernizing the data infrastructure in a way where you can answer questions like that, where you can you know, pull in data from any particular rep to process to make available for search.[00:12:34] And, and moving all of that data into a format where you can do all of this in minutes as opposed to, you know, days or weeks or months. That laid a lot of the groundwork for building anything in, in ai, at least in terms of training our own own models and then fine tuning them with, with replica data.[00:12:50] So then you know, we, we started a team last year recruited people from, you know from a team of, of zero or a team of one to, to the AI and data team today. We, we build. Everything related to, to ghostrider. So that means the various features like explain code, generate code, transform Code, and Ghostrider chat which is like a in context ide or a chat product within the, in the ide.[00:13:18] And then the code completion models, which are ghostwriter code complete, which was the, the very first version of, of ghostrider. Yeah. And we also support, you know, things like search and, and anything in terms of what creates, or anything that requires like large data scale or large scale processing of, of data for the site.[00:13:38] And, and various types of like ML algorithms for the site, for internal use of the site to do things like detect and stop abuse. Mm-hmm.[00:13:47] Alessio Fanelli: Yep. Sounds like a lot of the early stuff you worked on was more analytical, kind of like analyzing data, getting answers on these things. Obviously this has evolved now into some.[00:13:57] Production use case code lms, how is the team? And maybe like some of the skills changed. I know there's a lot of people wondering, oh, I was like a modern data stack expert, or whatever. It's like I was doing feature development, like, how's my job gonna change? Like,[00:14:12] Reza Shabani: yeah. It's a good question. I mean, I think that with with language models, the shift has kind of been from, or from traditional ml, a lot of the shift has gone towards more like nlp backed ml, I guess.[00:14:26] And so, you know, there, there's an entire skill set of applicants that I no longer see, at least for, for this role which are like people who know how to do time series and, and ML across time. Right. And, and you, yeah. Like you, you know, that exact feeling of how difficult it is to. You know, you have like some, some text or some variable and then all of a sudden you wanna track that over time.[00:14:50] The number of dimensions that it, that it introduces is just wild and it's a totally different skill set than what we do in a, for example, in in language models. And it's very it's a, it's a skill that is kind of you know, at, at least at rep not used much. And I'm sure in other places used a lot, but a lot of the, the kind of excitement about language models has pulled away attention from some of these other ML areas, which are extremely important and, and I think still going to be valuable.[00:15:21] So I would just recommend like anyone who is a, a data stack expert, like of course it's cool to work with NLP and text data and whatnot, but I do think at some point it's going to you know, having, having skills outside of that area and in more traditional aspects of ML will, will certainly be valuable as well.[00:15:39] swyx: Yeah. I, I'd like to spend a little bit of time on this data stack notion pitch. You were even, you were effectively the first data hire at rep. And I just spent the past year myself diving into data ecosystem. I think a lot of software engineers are actually. Completely unaware that basically every company now eventually evolves.[00:15:57] The data team and the data team does everything that you just mentioned. Yeah. All of us do exactly the same things, set up the same pipelines you know, shop at the same warehouses essentially. Yeah, yeah, yeah, yeah. So that they enable everyone else to query whatever they, whatever they want. And to, to find those insights that that can drive their business.[00:16:15] Because everyone wants to be data driven. They don't want to do the janitorial work that it comes, that comes to, yeah. Yeah. Hooking everything up. What like, so rep is that you think like 90 ish people now, and then you, you joined two years ago. Was it like 30 ish people? Yeah, exactly. We're 30 people where I joined.[00:16:30] So and I just wanna establish your founders. That is exactly when we hired our first data hire at Vilify as well. I think this is just a very common pattern that most founders should be aware of, that like, You start to build a data discipline at this point. And it's, and by the way, a lot of ex finance people very good at this because that's what we do at our finance job.[00:16:48] Reza Shabani: Yeah. Yeah. I was, I was actually gonna Good say that is that in, in some ways, you're kind of like the perfect first data hire because it, you know, you know how to build things in a reliable but fast way and, and how to build them in a way that, you know, it's, it scales over time and evolves over time because financial markets move so quickly that if you were to take all of your time building up these massive systems, like the trading opportunities gone.[00:17:14] So, yeah. Yeah, they're very good at it. Cool. Okay. Well,[00:17:18] swyx: I wanted to cover Ghost Writer as a standalone thing first. Okay. Yeah. And then go into code, you know, V1 or whatever you're calling it. Yeah. Okay. Okay. That sounds good. So order it[00:17:26] Replit GhostWriter[00:17:26] Reza Shabani: however you like. Sure. So the original version of, of Ghost Writer we shipped in August of, of last year.[00:17:33] Yeah. And so this was a. This was a code completion model similar to GitHub's co-pilot. And so, you know, you would have some text and then it would predict like, what, what comes next. And this was, the original version was actually based off of the cogen model. And so this was an open source model developed by Salesforce that was trained on, on tons of publicly available code data.[00:17:58] And so then we took their their model, one of the smaller ones, did some distillation some other kind of fancy tricks to, to make it much faster and and deployed that. And so the innovation there was really around how to reduce the model footprint in a, to, to a size where we could actually serve it to, to our users.[00:18:20] And so the original Ghost Rider You know, we leaned heavily on, on open source. And our, our friends at Salesforce obviously were huge in that, in, in developing these models. And, but, but it was game changing just because we were the first startup to actually put something like that into production.[00:18:38] And, and at the time, you know, if you wanted something like that, there was only one, one name and, and one place in town to, to get it. And and at the same time, I think I, I'm not sure if that's like when the image models were also becoming open sourced for the first time. And so the world went from this place where, you know, there was like literally one company that had all of these, these really advanced models to, oh wait, maybe these things will be everywhere.[00:19:04] And that's exactly what's happened in, in the last Year or so, as, as the models get more powerful and then you always kind of see like an open source version come out that someone else can, can build and put into production very quickly at, at, you know, a fraction of, of the cost. So yeah, that was the, the kind of code completion Go Strider was, was really just, just that we wanted to fine tune it a lot to kind of change the way that our users could interact with it.[00:19:31] So just to make it you know, more customizable for our use cases on, on Rep. And so people on Relet write a lot of, like jsx for example, which I don't think was in the original training set for, for cogen. And and they do specific things that are more Tuned to like html, like they might wanna run, right?[00:19:50] Like inline style or like inline CSS basically. Those types of things. And so we experimented with fine tuning cogen a bit here and there, and, and the results just kind of weren't, weren't there, they weren't where you know, we, we wanted the model to be. And, and then we just figured we should just build our own infrastructure to, you know, train these things from scratch.[00:20:11] Like, LMS aren't going anywhere. This world's not, you know, it's, it's not like we're not going back to that world of there's just one, one game in town. And and we had the skills infrastructure and the, and the team to do it. So we just started doing that. And you know, we'll be this week releasing our very first open source code model.[00:20:31] And,[00:20:31] Benchmarking Code LLMs[00:20:31] Alessio Fanelli: and when you say it was not where you wanted it to be, how were you benchmarking[00:20:36] Reza Shabani: it? In that particular case, we were actually, so, so we have really two sets of benchmarks that, that we use. One is human eval, so just the standard kind of benchmark for, for Python, where you can generate some code or you give you give the model a function definition with, with some string describing what it's supposed to do, and then you allow it to complete that function, and then you run a unit test against it and and see if what it generated passes the test.[00:21:02] So we, we always kind of, we would run this on the, on the model. The, the funny thing is the fine tuned versions of. Of Cogen actually did pretty well on, on that benchmark. But then when we, we then have something called instead of human eval. We call it Amjad eval, which is basically like, what does Amjad think?[00:21:22] Yeah, it's, it's exactly that. It's like testing the vibes of, of a model. And it's, it's cra like I've never seen him, I, I've never seen anyone test the model so thoroughly in such a short amount of time. He's, he's like, he knows exactly what to write and, and how to prompt the model to, to get you know, a very quick read on, on its quote unquote vibes.[00:21:43] And and we take that like really seriously. And I, I remember there was like one, one time where we trained a model that had really good you know, human eval scores. And the vibes were just terrible. Like, it just wouldn't, you know, it, it seemed overtrained. So so that's a lot of what we found is like we, we just couldn't get it to Pass the vibes test no matter how the, how[00:22:04] swyx: eval.[00:22:04] Well, can you formalize I'm jal because I, I actually have a problem. Slight discomfort with human eval. Effectively being the only code benchmark Yeah. That we have. Yeah. Isn't that[00:22:14] Reza Shabani: weird? It's bizarre. It's, it's, it's weird that we can't do better than that in some, some way. So, okay. If[00:22:21] swyx: I, if I asked you to formalize Mja, what does he look for that human eval doesn't do well on?[00:22:25] Reza Shabani: Ah, that is a, that's a great question. A lot of it is kind of a lot of it is contextual like deep within, within specific functions. Let me think about this.[00:22:38] swyx: Yeah, we, we can pause for. And if you need to pull up something.[00:22:41] Reza Shabani: Yeah, I, let me, let me pull up a few. This, this[00:22:43] swyx: is gold, this catnip for people.[00:22:45] Okay. Because we might actually influence a benchmark being evolved, right. So, yeah. Yeah. That would be,[00:22:50] Reza Shabani: that would be huge. This was, this was his original message when he said the vibes test with, with flying colors. And so you have some, some ghostrider comparisons ghost Rider on the left, and cogen is on the right.[00:23:06] AmjadEval live demo[00:23:06] Reza Shabani: So here's Ghostrider. Okay.[00:23:09] swyx: So basically, so if I, if I summarize it from a, for ghosting the, there's a, there's a, there's a bunch of comments talking about how you basically implement a clone. Process or to to c Clooney process. And it's describing a bunch of possible states that he might want to, to match.[00:23:25] And then it asks for a single line of code for defining what possible values of a name space it might be to initialize it in amjadi val With what model is this? Is this your, this is model. This is the one we're releasing. Yeah. Yeah. It actually defines constants which are human readable and nice.[00:23:42] And then in the other cogen Salesforce model, it just initializes it to zero because it reads that it starts of an int Yeah, exactly. So[00:23:51] Reza Shabani: interesting. Yeah. So you had a much better explanation of, of that than than I did. It's okay. So this is, yeah. Handle operation. This is on the left.[00:24:00] Okay.[00:24:00] swyx: So this is rep's version. Yeah. Where it's implementing a function and an in filling, is that what it's doing inside of a sum operation?[00:24:07] Reza Shabani: This, so this one doesn't actually do the infill, so that's the completion inside of the, of the sum operation. But it, it's not, it's, it, it's not taking into account context after this value, but[00:24:18] swyx: Right, right.[00:24:19] So it's writing an inline lambda function in Python. Okay.[00:24:21] Reza Shabani: Mm-hmm. Versus[00:24:24] swyx: this one is just passing in the nearest available variable. It's, it can find, yeah.[00:24:30] Reza Shabani: Okay. So so, okay. I'll, I'll get some really good ones in a, in a second. So, okay. Here's tokenize. So[00:24:37] swyx: this is an assertion on a value, and it's helping to basically complete the entire, I think it looks like an E s T that you're writing here.[00:24:46] Mm-hmm. That's good. That that's, that's good. And then what does Salesforce cogen do? This is Salesforce cogen here. So is that invalidism way or what, what are we supposed to do? It's just making up tokens. Oh, okay. Yeah, yeah, yeah. So it's just, it's just much better at context. Yeah. Okay.[00:25:04] Reza Shabani: And, and I guess to be fair, we have to show a case where co cogen does better.[00:25:09] Okay. All right. So here's, here's one on the left right, which[00:25:12] swyx: is another assertion where it's just saying that if you pass in a list, it's going to throw an exception saying in an expectedly list and Salesforce code, Jen says,[00:25:24] Reza Shabani: This is so, so ghost writer was sure that the first argument needs to be a list[00:25:30] swyx: here.[00:25:30] So it hallucinated that it wanted a list. Yeah. Even though you never said it was gonna be a list.[00:25:35] Reza Shabani: Yeah. And it's, it's a argument of that. Yeah. Mm-hmm. So, okay, here's a, here's a cooler quiz for you all, cuz I struggled with this one for a second. Okay. What is.[00:25:47] swyx: Okay, so this is a four loop example from Amjad.[00:25:50] And it's, it's sort of like a q and a context in a chat bot. And it's, and it asks, and Amjad is asking, what does this code log? And it just paste in some JavaScript code. The JavaScript code is a four loop with a set time out inside of it with a cons. The console logs out the iteration variable of the for loop and increasing numbers of of, of times.[00:26:10] So it's, it goes from zero to five and then it just increases the, the delay between the timeouts each, each time. Yeah.[00:26:15] Reza Shabani: So, okay. So this answer was provided by by Bard. Mm-hmm. And does it look correct to you? Well,[00:26:22] the[00:26:22] Alessio Fanelli: numbers too, but it's not one second. It's the time between them increases.[00:26:27] It's like the first one, then the one is one second apart, then it's two seconds, three seconds. So[00:26:32] Reza Shabani: it's not, well, well, so I, you know, when I saw this and, and the, the message and the thread was like, Our model's better than Bard at, at coding Uhhuh. This is the Bard answer Uhhuh that looks totally right to me.[00:26:46] Yeah. And this is our[00:26:47] swyx: answer. It logs 5 5 55, what is it? Log five 50. 55 oh oh. Because because it logs the state of I, which is five by the time that the log happens. Mm-hmm. Yeah.[00:27:01] Reza Shabani: Oh God. So like we, you know we were shocked. Like, and, and the Bard dancer looked totally right to, to me. Yeah. And then, and somehow our code completion model mind Jude, like this is not a conversational chat model.[00:27:14] Mm-hmm. Somehow gets this right. And and, you know, Bard obviously a much larger much more capable model with all this fancy transfer learning and, and and whatnot. Some somehow, you know, doesn't get it right. So, This is the kind of stuff that goes into, into mja eval that you, you won't find in any benchmark.[00:27:35] Good. And and, and it's, it's the kind of thing that, you know, makes something pass a, a vibe test at Rep.[00:27:42] swyx: Okay. Well, okay, so me, this is not a vibe, this is not so much a vibe test as the, these are just interview questions. Yeah, that's, we're straight up just asking interview questions[00:27:50] Reza Shabani: right now. Yeah, no, the, the vibe test, the reason why it's really difficult to kind of show screenshots that have a vibe test is because it really kind of depends on like how snappy the completion is, how what the latency feels like and if it gets, if it, if it feels like it's making you more productive.[00:28:08] And and a lot of the time, you know, like the, the mix of, of really low latency and actually helpful content and, and helpful completions is what makes up the, the vibe test. And I think part of it is also, is it. Is it returning to you or the, the lack of it returning to you things that may look right, but be completely wrong.[00:28:30] I think that also kind of affects Yeah. Yeah. The, the vibe test as well. Yeah. And so, yeah, th this is very much like a, like a interview question. Yeah.[00:28:39] swyx: The, the one with the number of processes that, that was definitely a vibe test. Like what kind of code style do you expect in this situation? Yeah.[00:28:47] Is this another example? Okay.[00:28:49] Reza Shabani: Yeah. This is another example with some more Okay. Explanations.[00:28:53] swyx: Should we look at the Bard one[00:28:54] Reza Shabani: first? Sure. These are, I think these are, yeah. This is original GT three with full size 175. Billion[00:29:03] swyx: parameters. Okay, so you asked GPC three, I'm a highly intelligent question answering bot.[00:29:07] If you ask me a question that is rooted in truth, I'll give you the answer. If you ask me a question that is nonsense I will respond with unknown. And then you ask it a question. What is the square root of a bananas banana? It answers nine. So complete hallucination and failed to follow the instruction that you gave it.[00:29:22] I wonder if it follows if one, if you use an instruction to inversion it might, yeah. Do what better?[00:29:28] Reza Shabani: On, on the original[00:29:29] swyx: GP T Yeah, because I like it. Just, you're, you're giving an instructions and it's not[00:29:33] Reza Shabani: instruction tuned. Now. Now the interesting thing though is our model here, which does follow the instructions this is not instruction tuned yet, and we still are planning to instruction tune.[00:29:43] Right? So it's like for like, yeah, yeah, exactly. So,[00:29:45] swyx: So this is a replica model. Same question. What is the square of bananas? Banana. And it answers unknown. And this being one of the, the thing that Amjad was talking about, which you guys are. Finding as a discovery, which is, it's better on pure natural language questions, even though you trained it on code.[00:30:02] Exactly. Yeah. Hmm. Is that because there's a lot of comments in,[00:30:07] Reza Shabani: No. I mean, I think part of it is that there's a lot of comments and there's also a lot of natural language in, in a lot of code right. In terms of documentation, you know, you have a lot of like markdowns and restructured text and there's also just a lot of web-based code on, on replica, and HTML tends to have a lot of natural language in it.[00:30:27] But I don't think the comments from code would help it reason in this way. And, you know, where you can answer questions like based on instructions, for example. Okay. But yeah, it's, I know that that's like one of the things. That really shocked us is the kind of the, the fact that like, it's really good at, at natural language reasoning, even though it was trained on, on code.[00:30:49] swyx: Was this the reason that you started running your model on hella swag and[00:30:53] Reza Shabani: all the other Yeah, exactly. Interesting. And the, yeah, it's, it's kind of funny. Like it's in some ways it kind of makes sense. I mean, a lot of like code involves a lot of reasoning and logic which language models need and need to develop and, and whatnot.[00:31:09] And so you know, we, we have this hunch that maybe that using that as part of the training beforehand and then training it on natural language above and beyond that really tends to help. Yeah,[00:31:21] Aligning Models on Vibes[00:31:21] Alessio Fanelli: this is so interesting. I, I'm trying to think, how do you align a model on vibes? You know, like Bard, Bard is not purposefully being bad, right?[00:31:30] Like, there's obviously something either in like the training data, like how you're running the process that like, makes it so that the vibes are better. It's like when it, when it fails this test, like how do you go back to the team and say, Hey, we need to get better[00:31:44] Reza Shabani: vibes. Yeah, let's do, yeah. Yeah. It's a, it's a great question.[00:31:49] It's a di it's very difficult to do. It's not you know, so much of what goes into these models in, in the same way that we have no idea how we can get that question right. The programming you know, quiz question. Right. Whereas Bard got it wrong. We, we also have no idea how to take certain things out and or, and to, you know, remove certain aspects of, of vibes.[00:32:13] Of course there's, there's things you can do to like scrub the model, but it's, it's very difficult to, to get it to be better at something. It's, it's almost like all you can do is, is give it the right type of, of data that you think will do well. And then and, and of course later do some fancy type of like, instruction tuning or, or whatever else.[00:32:33] But a lot of what we do is finding the right mix of optimal data that we want to, to feed into the model and then hoping that the, that the data that's fed in is sufficiently representative of, of the type of generations that we want to do coming out. That's really the best that, that you can do.[00:32:51] Either the model has. Vibes or, or it doesn't, you can't teach vibes. Like you can't sprinkle additional vibes in it. Yeah, yeah, yeah. Same in real life. Yeah, exactly right. Yeah, exactly. You[00:33:04] Beyond Code Completion[00:33:04] Alessio Fanelli: mentioned, you know, co being the only show in town when you started, now you have this, there's obviously a, a bunch of them, right.[00:33:10] Cody, which we had on the podcast used to be Tap nine, kite, all these different, all these different things. Like, do you think the vibes are gonna be the main you know, way to differentiate them? Like, how are you thinking about. What's gonna make Ghost Rider, like stand apart or like, do you just expect this to be like table stakes for any tool?[00:33:28] So like, it just gonna be there?[00:33:30] Reza Shabani: Yeah. I, I do think it's, it's going to be table stakes for sure. I, I think that if you don't if you don't have AI assisted technology, especially in, in coding it's, it's just going to feel pretty antiquated. But but I do think that Ghost Rider stands apart from some of, of these other tools for for specific reasons too.[00:33:51] So this is kind of the, one of, one of the things that these models haven't really done yet is Come outside of code completion and outside of, of just a, a single editor file, right? So what they're doing is they're, they're predicting like the text that can come next, but they're not helping with the development process quite, quite yet outside of just completing code in a, in a text file.[00:34:16] And so the types of things that we wanna do with Ghost Rider are enable it to, to help in the software development process not just editing particular files. And so so that means using a right mix of like the right model for for the task at hand. But but we want Ghost Rider to be able to, to create scaffolding for you for, for these projects.[00:34:38] And so imagine if you would like Terraform. But, but powered by Ghostrider, right? I want to, I put up this website, I'm starting to get a ton of traffic to it and and maybe like I need to, to create a backend database. And so we want that to come from ghostrider as well, so it can actually look at your traffic, look at your code, and create.[00:34:59] You know a, a schema for you that you can then deploy in, in Postgres or, or whatever else? You know, I, I know like doing anything in in cloud can be a nightmare as well. Like if you wanna create a new service account and you wanna deploy you know, nodes on and, and have that service account, kind of talk to those nodes and return some, some other information, like those are the types of things that currently we have to kind of go, go back, go look at some documentation for Google Cloud, go look at how our code base does it you know, ask around in Slack, kind of figure that out and, and create a pull request.[00:35:31] Those are the types of things that we think we can automate away with with more advanced uses of, of ghostwriter once we go past, like, here's what would come next in, in this file. So, so that's the real promise of it, is, is the ability to help you kind of generate software instead of just code in a, in a particular file.[00:35:50] Ghostwriter Autonomous Agent[00:35:50] Reza Shabani: Are[00:35:50] Alessio Fanelli: you giving REPL access to the model? Like not rep, like the actual rep. Like once the model generates some of this code, especially when it's in the background, it's not, the completion use case can actually run the code to see if it works. There's like a cool open source project called Walgreen that does something like that.[00:36:07] It's like self-healing software. Like it gives a REPL access and like keeps running until it fixes[00:36:11] Reza Shabani: itself. Yeah. So, so, so right now there, so there's Ghostrider chat and Ghostrider code completion. So Ghostrider Chat does have, have that advantage in, in that it can it, it knows all the different parts of, of the ide and so for example, like if an error is thrown, it can look at the, the trace back and suggest like a fix for you.[00:36:33] So it has that type of integration. But the what, what we really want to do is is. Is merge the two in a way where we want Ghost Rider to be like, like an autonomous agent that can actually drive the ide. So in these action models, you know, where you have like a sequence of of events and then you can use you know, transformers to kind of keep track of that sequence and predict the next next event.[00:36:56] It's how, you know, companies like, like adapt work these like browser models that can, you know, go and scroll through different websites or, or take some, some series of actions in a, in a sequence. Well, it turns out the IDE is actually a perfect place to do that, right? So like when we talk about creating software, not just completing code in a file what do you do when you, when you build software?[00:37:17] You, you might clone a repo and then you, you know, will go and change some things. You might add a new file go down, highlight some text, delete that value, and point it to some new database, depending on the value in a different config file or in your environment. And then you would go in and add additional block code to, to extend its functionality and then you might deploy that.[00:37:40] Well, we, we have all of that data right there in the replica ide. And and we have like terabytes and terabytes of, of OT data you know, operational transform data. And so, you know, we can we can see that like this person has created a, a file what they call it, and, you know, they start typing in the file.[00:37:58] They go back and edit a different file to match the you know, the class name that they just put in, in the original file. All of that, that kind of sequence data is what we're looking to to train our next model on. And so that, that entire kind of process of actually building software within the I D E, not just like, here's some text what comes next, but rather the, the actions that go into, you know, creating a fully developed program.[00:38:25] And a lot of that includes, for example, like running the code and seeing does this work, does this do what I expected? Does it error out? And then what does it do in response to that error? So all, all of that is like, Insanely valuable information that we want to put into our, our next model. And and that's like, we think that one can be way more advanced than the, than this, you know, go straighter code completion model.[00:38:47] Releasing Replit-code-v1-3b[00:38:47] swyx: Cool. Well we wanted to dive in a little bit more on, on the model that you're releasing. Maybe we can just give people a high level what is being released what have you decided to open source and maybe why open source the story of the YOLO project and Yeah. I mean, it's a cool story and just tell it from the start.[00:39:06] Yeah.[00:39:06] Reza Shabani: So, so what's being released is the, the first version that we're going to release. It's a, it's a code model called replica Code V1 three B. So this is a relatively small model. It's 2.7 billion parameters. And it's a, it's the first llama style model for code. So, meaning it's just seen tons and tons of tokens.[00:39:26] It's been trained on 525 billion tokens of, of code all permissively licensed code. And it's it's three epox over the training set. And And, you know, all of that in a, in a 2.7 billion parameter model. And in addition to that, we, for, for this project or, and for this model, we trained our very own vocabulary as well.[00:39:48] So this, this doesn't use the cogen vocab. For, for the tokenize we, we trained a totally new tokenize on the underlying data from, from scratch, and we'll be open sourcing that as well. It has something like 32,000. The vocabulary size is, is in the 32 thousands as opposed to the 50 thousands.[00:40:08] Much more specific for, for code. And, and so it's smaller faster, that helps with inference, it helps with training and it can produce more relevant content just because of the you know, the, the vocab is very much trained on, on code as opposed to, to natural language. So, yeah, we'll be releasing that.[00:40:29] This week it'll be up on, on hugging pace so people can take it play with it, you know, fine tune it, do all type of things with it. We want to, we're eager and excited to see what people do with the, the code completion model. It's, it's small, it's very fast. We think it has great vibes, but we, we hope like other people feel the same way.[00:40:49] And yeah. And then after, after that, we might consider releasing the replica tuned model at, at some point as well, but still doing some, some more work around that.[00:40:58] swyx: Right? So there are actually two models, A replica code V1 three B and replica fine tune V1 three B. And the fine tune one is the one that has the 50% improvement in in common sense benchmarks, which is going from 20% to 30%.[00:41:13] For,[00:41:13] Reza Shabani: for yes. Yeah, yeah, yeah, exactly. And so, so that one, the, the additional tuning that was done on that was on the publicly available data on, on rep. And so, so that's, that's you know, data that's in public res is Permissively licensed. So fine tuning on on that. Then, Leads to a surprisingly better, like significantly better model, which is this retuned V1 three B, same size, you know, same, very fast inference, same vocabulary and everything.[00:41:46] The only difference is that it's been trained on additional replica data. Yeah.[00:41:50] swyx: And I think I'll call out that I think in one of the follow up q and as that Amjad mentioned, people had some concerns with using replica data. Not, I mean, the licensing is fine, it's more about the data quality because there's a lot of beginner code Yeah.[00:42:03] And a lot of maybe wrong code. Mm-hmm. But it apparently just wasn't an issue at all. You did[00:42:08] Reza Shabani: some filtering. Yeah. I mean, well, so, so we did some filtering, but, but as you know, it's when you're, when you're talking about data at that scale, it's impossible to keep out, you know, all of the, it's, it's impossible to find only select pieces of data that you want the, the model to see.[00:42:24] And, and so a lot of the, a lot of that kind of, you know, people who are learning to code material was in there anyway. And, and you know, we obviously did some quality filtering, but a lot of it went into the fine tuning process and it really helped for some reason. You know, there's a lot of high quality code on, on replica, but there's like you, like you said, a lot of beginner code as well.[00:42:46] And that was, that was the really surprising thing is that That somehow really improved the model and its reasoning capabilities. It felt much more kind of instruction tuned afterward. And, and you know, we have our kind of suspicions as as to why there's, there's a lot of like assignments on rep that kind of explain this is how you do something and then you might have like answers and, and whatnot.[00:43:06] There's a lot of people who learn to code on, on rep, right? And, and like, think of a beginner coder, like think of a code model that's learning to, to code learning this reasoning and logic. It's probably a lot more valuable to see that type of, you know, the, the type of stuff that you find on rep as opposed to like a large legacy code base that that is, you know, difficult to, to parse and, and figure out.[00:43:29] So, so that was very surprising to see, you know, just such a huge jump in in reasoning ability once trained on, on replica data.[00:43:38] The YOLO training run[00:43:38] swyx: Yeah. Perfect. So we're gonna do a little bit of storytelling just leading up to the, the an the developer day that you had last week. Yeah. My understanding is you decide, you raised some money, you decided to have a developer day, you had a bunch of announcements queued up.[00:43:52] And then you were like, let's train the language model. Yeah. You published a blog post and then you announced it on Devrel Day. What, what, and, and you called it the yolo, right? So like, let's just take us through like the[00:44:01] Reza Shabani: sequence of events. So so we had been building the infrastructure to kind of to, to be able to train our own models for, for months now.[00:44:08] And so that involves like laying out the infrastructure, being able to pull in the, the data processes at scale. Being able to do things like train your own tokenizes. And and even before this you know, we had to build out a lot of this data infrastructure for, for powering things like search.[00:44:24] There's over, I think the public number is like 200 and and 30 million res on, on re. And each of these res have like many different files and, and lots of code, lots of content. And so you can imagine like what it must be like to, to be able to query that, that amount of, of data in a, in a reasonable amount of time.[00:44:45] So we've You know, we spent a lot of time just building the infrastructure that allows for for us to do something like that and, and really optimize that. And, and this was by the end of last year. That was the case. Like I think I did a demo where I showed you can, you can go through all of replica data and parse the function signature of every Python function in like under two minutes.[00:45:07] And, and there's, you know, many, many of them. And so a and, and then leading up to developer day, you know, we had, we'd kind of set up these pipelines. We'd started training these, these models, deploying them into production, kind of iterating and, and getting that model training to production loop.[00:45:24] But we'd only really done like 1.3 billion parameter models. It was like all JavaScript or all Python. So there were still some things like we couldn't figure out like the most optimal way to to, to do it. So things like how do you pad or yeah, how do you how do you prefix chunks when you have like multi-language models, what's like the optimal way to do it and, and so on.[00:45:46] So you know, there's two PhDs on, on the team. Myself and Mike and PhDs tend to be like careful about, you know, a systematic approach and, and whatnot. And so we had this whole like list of things we were gonna do, like, oh, we'll test it on this thing and, and so on. And even these, like 1.3 billion parameter models, they were only trained on maybe like 20 billion tokens or 30 billion tokens.[00:46:10] And and then Amjad joins the call and he's like, no, let's just, let's just yolo this. Like, let's just, you know, we're raising money. Like we should have a better code model. Like, let's yolo it. Let's like run it on all the data. How many tokens do we have? And, and, and we're like, you know, both Michael and I are like, I, I looked at 'em during the call and we were both like, oh God is like, are we really just gonna do this?[00:46:33] And[00:46:34] swyx: well, what is the what's the hangup? I mean, you know that large models work,[00:46:37] Reza Shabani: you know that they work, but you, you also don't know whether or not you can improve the process in, in In important ways by doing more data work, scrubbing additional content, and, and also it's expensive. It's like, it, it can, you know it can cost quite a bit and if you, and if you do it incorrectly, you can actually get it.[00:47:00] Or you, you know, it's[00:47:02] swyx: like you hit button, the button, the go button once and you sit, sit back for three days.[00:47:05] Reza Shabani: Exactly. Yeah. Right. Well, like more like two days. Yeah. Well, in, in our case, yeah, two days if you're running 256 GP 100. Yeah. Yeah. And and, and then when that comes back, you know, you have to take some time to kind of to test it.[00:47:19] And then if it fails and you can't really figure out why, and like, yeah, it's, it's just a, it's kind of like a, a. A time consuming process and you just don't know what's going to, to come out of it. But no, I mean, I'm Judd was like, no, let's just train it on all the data. How many tokens do we have? We tell him and he is like, that's not enough.[00:47:38] Where can we get more tokens? Okay. And so Michele had this you know, great idea to to train it on multiple epox and so[00:47:45] swyx: resampling the same data again.[00:47:47] Reza Shabani: Yeah. Which, which can be, which is known risky or like, or tends to overfit. Yeah, you can, you can over overfit. But you know, he, he pointed us to some evidence that actually maybe this isn't really a going to be a problem.[00:48:00] And, and he was very persuasive in, in doing that. And so it, it was risky and, and you know, we did that training. It turned out. Like to actually be great for that, for that base model. And so then we decided like, let's keep pushing. We have 256 TVs running. Let's see what else we can do with it.[00:48:20] So we ran a couple other implementations. We ran you know, a the fine tune version as I, as I said, and that's where it becomes really valuable to have had that entire pipeline built out because then we can pull all the right data, de-dupe it, like go through the, the entire like processing stack that we had done for like months.[00:48:41] We did that in, in a matter of like two days for, for the replica data as well removed, you know, any of, any personal any pii like personal information removed, harmful content, removed, any of, of that stuff. And we just put it back through the that same pipeline and then trained on top of that.[00:48:59] And so I believe that replica tune data has seen something like 680. Billion tokens. And, and that's in terms of code, I mean, that's like a, a universe of code. There really isn't that much more out there. And, and it, you know, gave us really, really promising results. And then we also did like a UL two run, which allows like fill the middle capabilities and and, and will be, you know working to deploy that on, on rep and test that out as well soon.[00:49:29] But it was really just one of those Those cases where, like, leading up to developer day, had we, had we done this in this more like careful, systematic way what, what would've occurred in probably like two, three months. I got us to do it in, in a week. That's fun. It was a lot of fun. Yeah.[00:49:49] Scaling Laws: from Kaplan to Chinchilla to LLaMA[00:49:49] Alessio Fanelli: And so every time I, I've seen the stable releases to every time none of these models fit, like the chinchilla loss in, in quotes, which is supposed to be, you know, 20 tokens per, per, what's this part of the yo run?[00:50:04] Or like, you're just like, let's just throw out the tokens at it doesn't matter. What's most efficient or like, do you think there's something about some of these scaling laws where like, yeah, maybe it's good in theory, but I'd rather not risk it and just throw out the tokens that I have at it? Yeah,[00:50:18] Reza Shabani: I think it's, it's hard to, it's hard to tell just because there's.[00:50:23] You know, like, like I said, like these runs are expensive and they haven't, if, if you think about how many, how often these runs have been done, like the number of models out there and then, and then thoroughly tested in some forum. And, and so I don't mean just like human eval, but actually in front of actual users for actual inference as part of a, a real product that, that people are using.[00:50:45] I mean, it's not that many. And, and so it's not like there's there's like really well established kind of rules as to whether or not something like that could lead to, to crazy amounts of overfitting or not. You just kind of have to use some, some intuition around it. And, and what we kind of found is that our, our results seem to imply that we've really been under training these, these models.[00:51:06] Oh my god. And so like that, you know, all, all of the compute that we kind of. Through, with this and, and the number of tokens, it, it really seems to help and really seems to to improve. And I, and I think, you know, these things kind of happen where in, in the literature where everyone kind of converges to something seems to take it for for a fact.[00:51:27] And like, like Chinchilla is a great example of like, okay, you know, 20 tokens. Yeah. And but, but then, you know, until someone else comes along and kind of tries tries it out and sees actually this seems to work better. And then from our results, it seems imply actually maybe even even lla. Maybe Undertrained.[00:51:45] And, and it may be better to go even You know, like train on on even more tokens then and for, for the[00:51:52] swyx: listener, like the original scaling law was Kaplan, which is 1.7. Mm-hmm. And then Chin established 20. Yeah. And now Lama style seems to mean 200 x tokens to parameters, ratio. Yeah. So obviously you should go to 2000 X, right?[00:52:06] Like, I mean, it's,[00:52:08] Reza Shabani: I mean, we're, we're kind of out of code at that point, you know, it's like there, there is a real shortage of it, but I know that I, I know there are people working on I don't know if it's quite 2000, but it's, it's getting close on you know language models. And so our friends at at Mosaic are are working on some of these really, really big models that are, you know, language because you with just code, you, you end up running out of out of context.[00:52:31] So Jonathan at, at Mosaic has Jonathan and Naveen both have really interesting content on, on Twitter about that. Yeah. And I just highly recommend following Jonathan. Yeah,[00:52:43] MosaicML[00:52:43] swyx: I'm sure you do. Well, CAGR, can we talk about, so I, I was sitting next to Naveen. I'm sure he's very, very happy that you, you guys had such, such success with Mosaic.[00:52:50] Maybe could, could you shout out like what Mosaic did to help you out? What, what they do well, what maybe people don't appreciate about having a trusted infrastructure provider versus a commodity GPU provider?[00:53:01] Reza Shabani: Yeah, so I mean, I, I talked about this a little bit in the in, in the blog post in terms of like what, what advantages like Mosaic offers and, and you know, keep in mind, like we had, we had deployed our own training infrastructure before this, and so we had some experience with it.[00:53:15] It wasn't like we had just, just tried Mosaic And, and some of those things. One is like you can actually get GPUs from different providers and you don't need to be you know, signed up for that cloud provider. So it's, it kind of detaches like your GPU offering from the rest of your cloud because most of our cloud runs in, in gcp.[00:53:34] But you know, this allowed us to leverage GPUs and other providers as well. And then another thing is like train or infrastructure as a service. So you know, these GPUs burn out. You have note failures, you have like all, all kinds of hardware issues that come up. And so the ability to kind of not have to deal with that and, and allow mosaic and team to kind of provide that type of, of fault tolerance was huge for us.[00:53:59] As well as a lot of their preconfigured l m configurations for, for these runs. And so they have a lot of experience in, in training these models. And so they have. You know, the, the right kind of pre-configured setups for, for various models that make sure that, you know, you have the right learning rates, the right training parameters, and that you're making the, the best use of the GPU and, and the underlying hardware.[00:54:26] And so you know, your GPU utilization is always at, at optimal levels. You have like fewer law spikes than if you do, you can recover from them. And you're really getting the most value out of, out of the compute that you're kind of throwing at, at your data. We found that to be incredibly, incredibly helpful.[00:54:44] And so it, of the time that we spent running things on Mosaic, like very little of that time is trying to figure out why the G P U isn't being utilized or why you know, it keeps crashing or, or why we, you have like a cuda out of memory errors or something like that. So like all, all of those things that make training a nightmare Are are, you know, really well handled by, by Mosaic and the composer cloud and and ecosystem.[00:55:12] Yeah. I was gonna[00:55:13] swyx: ask cuz you're on gcp if you're attempted to rewrite things for the TPUs. Cause Google's always saying that it's more efficient and faster, whatever, but no one has experience with them. Yeah.[00:55:23] Reza Shabani: That's kind of the problem is that no one's building on them, right? Yeah. Like, like we want to build on, on systems that everyone else is, is building for.[00:55:31] Yeah. And and so with, with the, with the TPUs that it's not easy to do that.[00:55:36] Replit's Plans for the Future (and Hiring!)[00:55:36] swyx: So plans for the future, like hard problems that you wanna solve? Maybe like what, what do you like what kind of people that you're hiring on your team?[00:55:44] Reza Shabani: Yeah. So We are, we're currently hiring for for two different roles on, on my team.[00:55:49] Although we, you know, welcome applications from anyone that, that thinks they can contribute in, in this area. Replica tends to be like a, a band of misfits. And, and the type of people we work with and, and have on our team are you know, like just the, the perfect mix to, to do amazing projects like this with very, very few people.[00:56:09] Right now we're hiring for the applied a applied to AI ml engineer. And so, you know, this is someone who's. Creating data pipelines, processing the data at scale creating runs and and training models and you

What can data tell us when it comes to how our money is invested? Are there data science tools that can help us manage the ups and downs of the financial markets? How has machine learning impacted forecasting? Can we rely on AI for investment advice? On today's episode we explore these questions and more during a deep dive discussion on financial markets with our expert guest, Professor Andrew Lo. Our Guest: Andrew W. Lo is the Charles and Susan Harris Professor of Finance at the MIT Sloan School of Management, Director of the MIT Laboratory for Financial Engineering, and a principal investigator at the MIT Computer Science and Artificial Intelligence Laboratory. Professor Lo was recognized for his work on financial markets by being named one of TIMEMagazine's 100 most influential people in the world.

Funding new drug development is challenging. Trials are expensive, complex, and lengthy, and only a fraction of the therapeutics that go into clinical trials will ever come out the other side approved. Sarah Shores welcomes Dr. Andrew Lo, economist, professor at MIT and co-founder of QLS Advisors, a Quantitative Life Sciences Company working to accelerate the pace of change in biomedicine.Sources referenced: Stein, R. (2017) Incentivizing Charity: A New Way to Fund Cancer Research; Healthcare Finance: Modern Financial Analysis for Accelerating Biomedical Innovation by Andrew Lo, Shomesh E. ChaudhuriDiversification does not assure a profit and may not protect against loss of principal. Diversification among investment options and asset classes may help to reduce overall volatility. Risk management and due diligence processes seek to mitigate, but cannot eliminate, risk nor do they imply low risk.This material is intended for information purposes only, and does not constitute investment advice, a recommendation or an offer or solicitation to purchase or sell any securities, funds or strategies to any person in any jurisdiction in which an offer, solicitation, purchase or sale would be unlawful under the securities laws of such jurisdiction. The opinions expressed are as of the date of publication and are subject to change without notice. Reliance upon information in this material is at the sole discretion of the reader. Investing involves risks. BlackRock does and may seek to do business with companies covered in this podcast. As a result, readers should be aware that the firm may have a conflict of interest that could affect the objectivity of this podcast. In the U.S. and Canada, this material is intended for public distribution. In the UK and Non-European Economic Area (EEA) countries: this is Issued by BlackRock Investment Management (UK) Limited, authorised and regulated by the Financial Conduct Authority. Registered office: 12 Throgmorton Avenue, London, EC2N 2DL. Tel:+ 44 (0)20 7743 3000. Registered in England and Wales No. 02020394. For your protection telephone calls are usually recorded. Please refer to the Financial Conduct Authority website for a list of authorised activities conducted by BlackRock. In the European Economic Area (EEA): this is Issued by BlackRock (Netherlands) B.V. is authorised and regulated by the Netherlands Authority for the Financial Markets. Registered office Amstelplein 1, 1096 HA, Amsterdam, Tel: 020 – 549 5200, Tel: 31-20-549-5200. Trade Register No. 17068311 For your protection telephone calls are usually recorded.For Investors in Switzerland: This document is marketing material.In Singapore, this is issued by BlackRock (Singapore) Limited (Co. registration no. 200010143N). This advertisement or publication has not been reviewed by the Monetary Authority of Singapore. In Hong Kong, this material is issued by BlackRock MKTGH0222U/M-2036247-6/7Asset Management North Asia Limited and has not been reviewed by the Securities and Futures Commission of Hong Kong. In Australia, issued by BlackRock Investment Management (Australia) Limited ABN 13 006 165 975 AFSL 230 523 (BIMAL). The material provides general information only and does not take into account your individual objectives, financial situation, needs or circumstances. Refer to BIMAL's Financial Services Guide on its website for more information. Before making any investment decision, you should assess whether the material is appropriate for you and obtain financial advice tailored to you having regard to your individual objectives, financial situation, needs and circumstances.In Latin America: this material is for educational purposes only and does not constitute investment advice nor an offer or solicitation to sell or a solicitation of an offer to buy any shares of any Fund (nor shall any such shares be offered or sold to any person) in any jurisdiction in which an offer, solicitation, purchase or sale would be unlawful under the securities law of that jurisdiction. If any funds are mentioned or inferred to in this material, it is possible that some or all of the funds may not have been registered with the securities regulator of Argentina, Brazil, Chile, Colombia, Mexico, Panama, Peru, Uruguay or any other securities regulator in any Latin American country and thus might not be publicly offered within any such country. The securities regulators of such countries have not confirmed the accuracy of any information contained herein. The provision of investment management and investment advisory services is a regulated activity in Mexico thus is subject to strict rules. For more information on the Investment Advisory Services offered by BlackRock Mexico please refer to the Investment Services Guide available at www.blackrock.com/mx©2022 BlackRock, Inc. All Rights Reserved. BLACKROCK is a registered trademark ofBlackRock, Inc. All other trademarks are those of their respective owners.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

My guest today is Steve Foerster. Steve is a Professor of Finance at Ivey Business School and holds a Ph.D. from the Wharton School and a CFA designation. He's won numerous teaching and research awards, including the William F. Sharpe Award, written textbooks, and is a former director and chair of the investment committee of Western's alumni endowment fund. On top of all of that, he's published over 50 articles in journals such as the Journal of Financial Economics, the Journal of Finance, and the Financial Analysts Journal. His list of accomplishments are incredible, and I'm thrilled to have him on the show today to dig into his wisdom. In his latest book, In Pursuit of the Perfect Portfolio: The Stories, Voices, and Key Insights of the Pioneers Who Shaped the Way We Invest, Steve and coauthor Andrew Lo tell the story of modern investing by profiling and interviewing some of the most prominent figures in the world of finance, including six Nobel Laureates and an innovator in the world of mutual funds. Through this work, he allows us to explore what the perfect portfolio might be–and shares fascinating information about how we invest today. In our conversation, Steve and I discussed how he chose the 10 innovative investors who became the focus of his new book, how to take distinct investing approaches, and create a “super portfolio.” We'll also dig into what goes wrong in so many conversations between investors of all ages and their financial advisors and much more. GET A FREE COPY OF STEVE'S BOOK, IN PURSUIT OF THE PERFECT PORTFOLIO: THE STORIES, VOICES, AND KEY INSIGHTS OF THE PIONEERS WHO SHAPED THE WAY WE INVEST. Here's all you have to do... Step 1.) Subscribe to the podcast and leave an honest rating & review over on iTunes. Step 2.) Text BOOK, that's B-O-O-K to 866-482-9559 for a link to our book request page, complete the form and we will ship you the book for free. It's that simple! Show Notes: RetireWithPurpose.com/327 Rate & Review the Podcast: RetireWithPurpose.com/review Sign Up to Casey's Weekend Reading Email! Sifting through the copious amount of conflicting financial advice and retirement information can be daunting - but it doesn't have to be! Each week, Casey makes it super easy. He hand-picks 4 of the most important articles you need to read, that are beneficial to you whether you're at, near, or in retirement! If you want them sent straight to your inbox, sign up by visiting RetireWithPurpose.com/weekend-reading

It's a killer that's been swept under the rug. The leading cause of death for 18–45-year-olds in America is opioid overdose, and deaths are only increasing. Liberty and Scott want to know, why is this decades-old epidemic still taking so many lives? They're getting the facts on the opioid crisis, looking at the local community and financial perspectives to get the truth. They go to the experts to explore the local community perspective and the financial aspect to discover the solutions that can stop more Americans from dying of opioid overdoses. They speak with Andrew Lo, professor of finance at the MIT Sloan School of Management; and with Chief Tom Synan, chief of police in Ohio.  Data Nation is a production of the MIT Institute for Data, Systems, and Society and Voxtopica. 

In this episode, Ed sits down Process Programming coaches Liam Carmichael, Lotte Thompson and Andrew Lo, to talk about the 2022 CrossFit Open and Quarterfinals. As a group they reflect on their personal experiences from the years competitions: Mistakes made, lessons learned, what would they change if they had another chance and their goals for the next season.   Sponsors: The Process Programming: Website: www.theprocessprogramming.com Instagram:  @theprocessprogramming

Gonzalo, gran aficionado a las series, disfruta con las lecciones de Jerry Seinfeld y Larry David. Los cómicos explican, mejor que los académicos, este mundo loco en el que vivimos. Si de él dependiera, The economics of Seinfeld convalidaría el primer año de ICADE. La economía es la ciencia que estudia la gestión de la escasez y los modelos permiten tomar mejores decisiones personales. Desde la elección de un trabajo con pocos ofertantes hasta la búsqueda en Tinder con información asimétrica.Escucha el podcast en tu plataforma habitual:Spotify — Apple — iVoox — YouTubeArtículos sobre finanzas en formato blog:Substack Kapital — Substack CardinalApuntes:Todo se puede entrenar. Toni Nadal.La imprescindible escuela de la dificultad. Toni Nadal.The market for lemons. George Akerlof.El penalti de Nash. Ignacio Palacios-Huerta.An economist's lessons on happiness. Richard Easterlin.Sexual conflict in human mating. David Buss.Psychological sex differences. David Buss.Adaptative markets. Andrew Lo.Freakonomics. Steven Levitt & Stephen Dubner.Superfreakonomics. Steven Levitt & Stephen Dubner.0.28. La oferta y la demanda en el mercado matrimonial.20.50. Dr. Strangelove, la bomba atómica en la teoría de juegos.31.44. «No tengo opiniones. Tengo una opinión: uno deber evaluar los datos».35.37. «No creo en la dureza como fin sino que creo en la dureza como medio».44.01. «We salute the rank, not the man».1.00.58. El coste de oportunidad en la elección de una carrera.1.12.26. La ventaja absoluta y la ventaja comparativa.1.22.55. El incremento marginal, en la unidad extra.1.32.45. La revolución de la economía freak.

Our guest this week is Dr. Andrew Lo. Dr. Lo is the Charles E. & Susan T. Harris Professor, a professor of finance, and the director of the Laboratory for Financial Engineering at the MIT Sloan School of Management. His current research spans five areas, including evolutionary models of investor behavior and adaptive markets, systemic risk, and financial regulation, among others. Dr. Lo has published extensively in academic journals and authored a number of books including In Pursuit of the Perfect Portfolio, which he cowrote with Stephen Foerster. He has received numerous awards for his work and contributions to modern finance research throughout his career. He holds a bachelor's in economics from Yale University and an AM and Ph.D. in economics from Harvard University.BackgroundBioIn Pursuit of the Perfect Portfolio: The Stories, Voices, and Key Insights of the Pioneers Who Shaped the Way We Invest, by Andrew W. Lo and Stephen R. FoersterAdaptive Markets: Financial Evolution at the Speed of Thought, by Andrew W. LoHistory“Thirty Maidens of Geneva,” the Tontine Coffee-House, thetch.blog.com, Aug. 5, 2019.“Why 18th Century Swiss Bankers Bet on the Lives of Young Girls,” by Stephen Foerster, sfoerster-5338.medium.com, Sept. 2, 2021.John Maynard KeynesBenjamin GrahamHarry MarkowitzHarry MarkowitzModern Portfolio TheoryWhat Is a Gunslinger? William F. SharpeWilliam F. SharpeWhat Is the Sharpe Ratio?Capital Asset Pricing Model (CAPM)“Keynes the Stock Market Investor: A Quantitative Analysis,” by David Chambers, Elroy Dimson, and Justin Foo, papers.ssrn.com, Sept. 26, 2013.Eugene F. FamaEugene FamaWhat Is the Efficient Market Hypothesis?“Algorithmic Models of Investor Behavior,” by Andrew Lo and Alexander Remorov, eqderivatives.com, 2021.“In Pursuit of the Perfect Portfolio: Eugene Fama,” Interview with Andrew Lo and Eugene Fama, youtube.com, Dec. 15, 2016.“Why Artificial Intelligence May Not Be as Useful or as Challenging as Artificial Stupidity,” by Andrew Lo, hdsr.mitpress.mit.edu, July 1, 2019.John C. Bogle John Bogle Cost Matters HypothesisCharles D. EllisCharley EllisGreenwich Associates“Charley Ellis: Why Active Investing Is Still a Loser's Game,” The Long View podcast, Morningstar.com, May 27, 2020.Other“7 Principles to Help You Create Your Perfect Portfolio,” by Robert Powell, marketwatch.com, Nov. 10, 2021.

In this episode we answer emails from Jeff (x2), Popeye, and the anonymous Mycontactinfo (x2).  We discuss some feeble attempts at disparagement by former financial media people and put them in context, an essay on taking your money off the table and weathering financial downturns by Bill Bernstein and an free course from MIT.And THEN we our go through our weekly portfolio reviews of the seven sample portfolios you can find at Portfolios | Risk Parity Radio.  Additional links:The Bernstein Essay:  A Day to Remember - HumbleDollarThe MIT Course with Andrew Lo:  Course | Adaptive Markets: Financial Market Dynamics and Human Behavior | edXSupport the show (https://www.riskparityradio.com/support)

Investors are constantly searching for the perfect portfolio. Although it may be elusive, there are common principles that can allow all of us to get closer to it. We speak to Stephen Foerster, co-author of the new book "In Pursuit of the Perfect Portfolio: The Stories, Voices, and Key Insights of the Pioneers Who Shaped the Way We Invest" about some of those principles, and what some of history's best investors and researchers can teach us about them. Along with his co-author Andrew Lo, Stephen spoke to investing legends like Harry Markowitz, William Sharpe, Jeremy Siegel, Jack Bogle, Charley Ellis and Robert Shiller to capture their ideas on the construction of a perfect portfolio. We go through each of those interviews to identify the key lessons investors can learn from them. We hope you enjoy the discussion. ABOUT THE PODCAST Excess Returns is an investing podcast hosted by Jack Forehand (@practicalquant) and Justin Carbonneau (@jjcarbonneau), partners at Validea. Justin and Jack discuss a wide range of investing topics including factor investing, value investing, momentum investing, multi-factor investing, trend following, market valuation and more with the goal of helping those who watch and listen become better long term investors. SEE LATEST EPISODES https://www.validea.com/excess-returns-podcast FIND OUT MORE ABOUT VALIDEA https://www.validea.com FOLLOW OUR BLOG https://blog.validea.com FIND OUT MORE ABOUT VALIDEA CAPITAL https://www.valideacapital.com FOLLOW JACK Twitter: https://twitter.com/practicalquant LinkedIn: https://www.linkedin.com/in/jack-forehand-8015094 FOLLOW JUSTIN Twitter: https://twitter.com/jjcarbonneau LinkedIn: https://www.linkedin.com/in/jcarbonneau

'When you look into history, so-called derivative investment products that you think of as being recent, actually go back to 2400 BCE, in various forms. Call options go back to 600 BCE at least. In the 18th century BCE there were personal loans, as well as a liquid secondary market for these promissory notes. So what we think of as new inventions are actually very old.' - Steve FoersterIf we could gather all of the famous investing pioneers from world history into one room and ask them to build the perfect portfolio, what would it look like?  Well Steve Foerster and his co-author (and previous guest on the show) Andrew Lo, set out to do just that, with their new book called 'In Pursuit of the Perfect Portfolio'. I thought I'd invite Steve onto the show to discuss his new book, go through some of the 'golden threads' of investing that he came across, his journey in the world of finance, and of course, get his opinion on how close Trend Following is to 'the Perfect Portfolio'. Thank you for listening and please welcome to the show, our guest, Steve Foerster. In This Episode, You'll Learn: Steve's journey to becoming a Professor of Finance and how he ended up co-authoring a book with Professor Andrew Lo Some of the concepts from the legendary financial experts they interviewed for the book How closely linked the academic world of finance really is About some of the most important findings in investment research over the decades About the "untold" story of an early 20th-century mathematician, Louis Bachelier Harry Markowitz' story What Eugene Fama's Perfect Portfolio looks like About William Sharpe and his now famous 'Sharpe Ratio' ‘The missing link was that correlations are critical as well. In fact, as we know, the more securities that you put into a portfolio, it's those correlations or that co-variant component that really dominates.' - Steve FoersterHow the phrase 'beta' came about How computers enabled researchers to access deeper insights into the world of investing About the debate between active and passive investing Whether volatility should still be considered as a measure of risk About the progression of volatility data into today's investment models About the story of Eugene Fama How Steve views the question of 'what is the perfect portfolio?' Whether Steve considers Trend Following investing as, essentially, a perfect portfolio Follow Niels on https://twitter.com/toptraderslive (Twitter), https://www.linkedin.com/in/nielskaastruplarsen (LinkedIn), https://www.youtube.com/user/toptraderslive (YouTube) or via the https://www.toptradersunplugged.com/ (TTU website). IT's TRUE

What is the economy?  People used to tell stories about the exchange of goods and services in terms of flows and processes — but over the last few hundred years, economic theory veered toward measuring discrete amounts of objects.  Why?  The change has less to do with the objective nature of economies and more to do with what tools theorists had available.  And scientific instruments — be they material technologies or concepts — don't just make new things visible, but also hide things in new blind spots.  For instance, algebra does very well with ratios and quantities…but fails to properly address what markets do: how innovation works, where value comes from, and how economic actors navigate (and change) a fundamentally uncertain shifting landscape.  With the advent of computers, new opportunities emerge to study that which cannot be contained in an equation. Using algorithms, scientists can formalize complex behaviors – and thinking economics in both nouns and verbs provides a more complete and useful stereoscopic view of what we are and do.This week we speak with W. Brian Arthur of The Santa Fe Institute, Stanford University, and Xerox PARC about his recent essay, “Economics in Nouns and Verbs.” In this first part of a two-part conversation, we explore how a mathematics of static objects fails to describe economies in motion — and how a process-based approach can fill gaps in our understanding.  If you can't wait two weeks for Part Two, dig through our archives for more Brian Arthur in episodes 13 and 14.If you value our research and communication efforts, please subscribe to Complexity Podcast wherever you prefer to listen, rate and review us at Apple Podcasts, and/or consider making a donation at santafe.edu/give. You can find numerous other ways to engage with us — including job openings for both SFI staff and postdoctoral researchers, as well as open online courses — at santafe.edu/engage.Join our Facebook discussion group to meet like minds and talk about each episode.Podcast theme music by Mitch Mignano.Follow us on social media:Twitter • YouTube • Facebook • Instagram • LinkedInRelated Reading & Listening:• “Economics in Nouns and Verbs” by W. Brian Arthur (pre-print)• @sfiscience Twitter thread excerpting “Economics in Nouns and Verbs”• “Mathematical languages shape our understanding of time in physics” by Nicolas Gisin for Nature Physics• “Introduction to PNAS special issue on evolutionary models of financial markets” by Simon Levin & Andrew Lo• “The Information Theory of Individuality” by David Krakauer et al. for Theory in Biosciences• “On Coronavirus, Crisis, and Creative Opportunity with David Krakauer” on Complexity Podcast• “The Erotics of Becoming: XENOGENESIS and The Thing” by Eric White for Science Fiction Studies• “New model shows how social networks could help generate economic phenomena like inequality & business cycles” by INET Oxford on research by J. Doyne Farmer

In a companion interview to his June 7 talk with Stanford's Michael Snyder, Harry speaks this week with Noosheen Hashemi, who—with Snyder—co-founded the personalized health startup January.ai in 2017. The company focuses on helping users understand how their bodies respond to different foods and activities, so they can make diet and exercise choices that help them avoid unhealthy spikes in blood glucose levels.January's smartphone app collects blood glucose levels from disposable devices called continuous glucose monitors (CGMs), as well as heart rate data from patients' Fitbits or Apple Watches. The app also makes it easier for users to log the food they eat, and see what impact each food has on their glucose levels. Once the app has enough data, January's machine learning algorithms can start predicting the effects of different foods and activities on blood glucose. It can then recommend meals and exercise that'll help users keep their blood glucose in a healthy target range. The goal isn't to prevent glucose spikes completely, but rather to prevent diabetes from emerging over the long term in people at risk for a cluster of serious conditions known metabolic syndrome. That could help individuals live longer, healthier lives. And at a population level it could save billions in healthcare costs.Please rate and review MoneyBall Medicine on Apple Podcasts! Here's how to do that from an iPhone, iPad, or iPod touch:• Launch the “Podcasts” app on your device. If you can't find this app, swipe all the way to the left on your home screen until you're on the Search page. Tap the search field at the top and type in “Podcasts.” Apple's Podcasts app should show up in the search results.• Tap the Podcasts app icon, and after it opens, tap the Search field at the top, or the little magnifying glass icon in the lower right corner.• Type MoneyBall Medicine into the search field and press the Search button.• In the search results, click on the MoneyBall Medicine logo.• On the next page, scroll down until you see the Ratings & Reviews section. Below that, you'll see five purple stars.• Tap the stars to rate the show.• Scroll down a little farther. You'll see a purple link saying “Write a Review.”• On the next screen, you'll see the stars again. You can tap them to leave a rating if you haven't already.• In the Title field, type a summary for your review.• In the Review field, type your review.• When you're finished, click Send.• That's it, you're done. Thanks!Full TranscriptHarry Glorikian: I'm Harry Glorikian, and this is MoneyBall Medicine, the interview podcast where we meet researchers, entrepreneurs, and physicians who are using the power of data to improve patient health and make healthcare delivery more efficient. You can think of each episode as a new chapter in the never-ending audio version of my 2017 book, “MoneyBall Medicine: Thriving in the New Data-Driven Healthcare Market.” If you like the show, please do us a favor and leave a rating and review at Apple Podcasts.Harry Glorikian: I've been making the show long enough that you can see a kind of family tree emerging, with branches that connect many of our episodes.That's definitely the case with today's interview with Noosheen Hashemi, the co-founder and CEO of the precision health company January AI.The branch leading to Hashemi started back in June of 2021 when I interviewed Professor Michael Snyder, the chair of Stanford's Department of Genetics.Snyder is a huge proponent of using wearable devices to help people make better decisions about their own health. In fact, the day we spoke he was wearing seven separate devices, including one called a continuous glucose monitor or CGM.A CGM is standard equipment these days for about 3.5 million diabetics in the U.S. who need to know when their blood sugar is too high and when it's time to take more insulin. But Snyder believes that blood glucose data could also help tens of millions of other people who don't yet take insulin but may be on their way to developing full-blown diabetes.Back in 2016 Snyder got a visit from Hashemi. She's a longtime Silicon Valley tech executive and philanthropist who'd been searching for a way to use AI, wearable devices, and big data to get more people involved in medical research. Hashemi told me it took just two meetings for her and Snyder to decide to join forces to co-found January. The company makes a smartphone app that collects blood glucose data from disposable CGMs, as well as heart rate data from patients' existing wearable devices such as their Fitbit or Apple Watch. The app also makes it easier for users to log the food they eat, and see what impact each food has on their glucose levels. Once the app has enough data, January's machine learning algorithms can start predicting the effects of different foods and activities on blood glucose. It can then recommend meals and exercise that'll help users keep their blood glucose in a healthy target range. The goal isn't to prevent glucose spikes completely, but rather to prevent diabetes from emerging over the long term in people at risk for a cluster of serious conditions known metabolic syndrome. That could help individuals live longer, healthier lives. And at a population level it could save billions in healthcare costs.As you're about to hear, Hashemi and I talked about why glucose monitoring is so important and what companies like January can do in the future to make the predictive power of AI available to more people.Harry Glorikian: Noosheen, welcome to the show. Noosheen Hashemi: Thank you, Harry. Harry Glorikian: So, it's great to have you on the show. It was interesting that, you know, the minute Dr. Snyder mentioned the company, I was immediately Googling it. And I was like, oh, I have to talk to this company. I have to understand what they're doing and, and what's going on.And to be quite honest, I've been doing my homework for the past couple of weeks. And I'm like: I think I have to call my doctor and get a ‘script to actually use the product. Just to help everybody get up to speed on this, can you bring people up to speed on where we are with glucose monitoring and health in general? Whether they have diabetes or whether they're just, you know, what, I, maybe someone like me who I hope is a generally a healthy person.Noosheen Hashemi: Sure, absolutely.  Yeah. So from Mike Snyder's four-year multi-omic IPOP research,  we learned that people who are so-called healthy and have healthy A1C levels could actually have huge glycemic variability. He sometimes calls these people with pre pre-diabetes.  I think eight people developed diabetes during his four-year study.There haven't been enough longitudinal studies in healthy people with glycemic variability to suggest that they will necessarily develop diabetes. So to date, there's really no conclusive evidence that healthy people can benefit from balancing their blood sugar. Also, not all sugar spikes are bad and a two-hour bike ride might produce a big spike, but that's fine. It's not the spike by itself that we worry about. It's really  how high the spike is against our baseline, against the population, whether the spike comes down quickly, the shape of the curve, the area under the curve. These are the things that are illuminating in terms of our state of metabolic health.So  at January we really view metabolic health as a spectrum. So we want to support people to figure out kind of where they are on that spectrum. And to try to really help them move up to healthier points on that spectrum. So we don't see it as a moment in time where you are something or you are not something. You are kind of on a spectrum of metabolic health, and we continuously want you to be self-aware and, and really improve your location on that spectrum. Now, something to keep in mind,  and why I think it's important for people to take action on this, is that 84% of the 88 million people believed to have pre-diabetes today, and 22% of the 34 million people that are believed to have diabetes today, are not diagnosed. They are undiagnosed. That's 75 million people walking around with pre-diabetes and don't even know. So, if we don't measure people's health, that doesn't mean they're healthy.  So we really encourage people to be  you know, vigilant with their health learn so that they can, they can act, you know, self-advocate. Be able to self-manage.So we do think that wearables are an easy, useful way to kind of see where things are, but then you need companies like January to make sense of it all. Harry Glorikian: Yeah. I mean  you know, it's interesting because you know, I'll go to my doctor and they'll do that one time measurement. It's like taking your car in and you're like, it was making a noise. It's not making the noise right now, but, you know, try and diagnose when that event is not happening. Whereas with the wearables, I can, I can actually see, you know, my, my heart rate variability change depending on my exercise process. I can see my sleep change if I had one too many glasses of wine. I have to tell you, I hate it because I would like to have more wine than my monitor allows me to have, but you know, you see the immediate feedback, which would let you sort of course-adjust accordingly. And you know, when I, there was a paper, I believe that was published in Israel where there, I think it was 500 people that they looked at and where you could see that every person, they could eat the same foods, but their spikes would be different or how long that spike would be based on genetics, based on their microbiome. And so if you're not monitoring, how will you know that your quote, healthy diet is actually healthy for you? Noosheen Hashemi: You don't.  You definitely don't. And yes, that's study shows variability between people, but also we've shown glycemic variability for the same person. So we had somebody at the office have the same good sleep nine days in a row, and they had a different glycemic response to that. Mostly every single day, nine days in a row, depending on how much they had slept, how stressed they were, how much workout they had done. And most importantly, how much fiber was in there.  So we are radically different person to person, and this is why we encourage people.  No one is going to know you as well as you do. And no one's going to be as interested in your health as you are  as you should be, as you might be. So we really encourage people to learn, learn, be self-aware self-advocate, self-educate. Harry Glorikian: So, help people understand this term metabolic syndrome, you know, and, and talk about how many people, maybe who are pre-diabetic go to full-blown diabetes, you know? Noosheen Hashemi: Okay. Yeah. So  I mentioned that 122 million people have either diabetes or pre-diabetes in America.  88 million plus 34 [million]. And then a larger number of people, if you believe Mike Snyder's pre-diabetes number, that's even a larger number. But metabolic syndrome is a cluster of conditions that leads to type 2 diabetes, heart disease, and stroke. These conditions are basically high blood sugar—which has been historically measured by A1C  blood tests called hemoglobin A1C, but increasingly it's measured by time and range using a CGM—high cholesterol and triglyceride levels, high blood pressure, high BMI, and high waist to hip ratio. So this kind of fat right in the middle.So the 2002 diabetes prevention study showed that unless there's an intervention, 58% of the people that have pre-diabetes could end up with diabetes. And usually they think of this prevention as weight loss.That's what the DPP programs, diabetes prevention programs, are about.So if you have pre-diabetes the cells in your body don't respond normally to insulin. And insulin is a hormone that facilitates your cells taking up glucose, which is a source of energy for your body. Your pancreas basically makes more insulin to try to get the cells to take up glucose. You sort of get into this terrible vicious circle. So eventually your pancreas can't keep up and then you have this sort of excess sugar sitting in your bloodstream, which is really a problem. And it can really lead to microvascular complications like retinopathy or neuropathy or diabetic nephropathy.So as you know, diabetic retinopathy is the most common cause of blindness in working adults in the developed world. And in diabetic neuropathy, essentially high blood sugar can injure nerves throughout the body. And usually damages nerves in the feet, in the legs and feet, which hear about foot ulcers and amputations coming from this.And of course  diabetic kidney disease. Nephropathy is something that  is the number one cause of kidney failure, actually. Almost a third of people with diabetes develop kidney disease. So you add this with the high blood pressure we can increase the force of blood through your arteries and damage arteries. And then you have excess blood pressure, you knowblood pressure and diabetes together, basically increase your risk for heart disease. So it's really a terrible cluster of conditions to have. And so if you have three of these conditions, three of these five, you essentially have metabolic syndrome. And if you have metabolic syndrome, you're at a higher risk of developing these different diseases. You really don't want to go down this path. The path itself is not great. And then the comorbidities from this path are just worse and complications of course are very painful, costly, and potentially, deadly.Harry Glorikian: And so that's one end of the spectrum, but in reality, even someone like me who tries to watch he eats, who goes running regularly, or tries to go running regularly. I mean, you know, I have sleep apnea because they tell me my BMI is too high. Right. So  but this sort of technology, you know, I could be spiking and keeping a high glucose level, which would inhibit my ability to lose weight, et cetera. So how can more data about blood glucose, and its relationship to diet, help people avoid diabetes?Noosheen Hashemi: Yeah. So for so long, we've been able, we've been told just to avoid refined sugar, refined flour, eat a lot of vegetables, walk 10,000 steps. You'll be fine. Or, you know, weight loss is given as the end goal to cure all diseases. You know, why don't you, Harry, drop 25 pounds? Or how about drop 5 to 10% of your weight? Harry Glorikian: Just like that!Noosheen Hashemi: It's true, weight loss really improves biomarkers. But how many people who get this advice can actually do that? And at the timeframe that they need to. So we feel like that's just not a practical approach to solving a problem.A more practical approach is to really figure out what works for each individual. You know, you mentioned you've dialed your own wine drinking based on its impact. I've done the same. I was, you know, enjoying two, three sips of wine. And then I learned that it would wake me up in the middle of the night. So I stopped having even the two, three sips of wine. So don't feel bad that you can't have your second and third and fourth glass. But basically we offer a multitude of levers that you can dial for your lifestyle. For example, intermittent fasting and calorie restriction together have shown benefits in clinical studies for improving insulin sensitivity, if you do them together. So you can't just fast and then gorge yourself. But if you fast and you restrict your calories together, you can really improve insulin sensitivity. So we let you, we help you using the January program to learn to experiment with fasting and calorie restriction and figure out what works for you. How much of it you can make. You know, slowly  help you essentially build it into your habits and your daily routines to fast. You know, we increase your fasting period 15 minutes at a time. So you may start with January you're eating 16 hours a day and you're fasting eight hours. You may end the program having reversed that.And other thing is we, we really pro promote fiber consumption. So increased fiber intake has been associated with higher levels of bacteria-derived short chain fatty acids, which is a regulator of GLP-1 production. As you know, GLP-1 is an incretin and a recognized regulator of glycemic homeostasis and satiety. So we help you track how much fiber you're eating. We encourage you to eat more, knowing what foods spike you, spike your blood sugar, helps you basically eliminate or reduce consumption of those foods. It tells you how much, how much of those things to eat  or alternatives that kind of honor your food preferences  and food tastes, but have lower glycemic index. If you can't walk 10,000 steps a day, okay. January tells you how much you need to walk, when you need to walk to keep your blood sugar in a healthy range. So you really need data  to, to dial your lifestyle. There are many levers and there are no silver bullets and there's too much to keep in your head. Which is why it's nice to have AI sort of help you kind of make, you know, take it all in to a platform and then synthesize it and give you insights.Harry Glorikian: Yeah. I mean, like,  I've got my, my Apple Watch. I've got my, you know, Whoop band. Right.I don't have as many as he [Mike Snyder] does, but  I know, I think my wife would kill me if I, if I was wearing eight things, but, but it's, you know, it's true. Like it's, you know, each one of these, because they're not holistically designed, give me a different piece of data that then I can then react to. You know, one is probably more of a coach that causes me to push a little bit farther, you know, et cetera. So  I mean, I hope one day we evolve to something that's a little bit more holistic so that the average person can sort of, it becomes more digestible and more actionable. But you know, I do believe, based on my conversation with him and even all the work that I do multi-factorial biomarkers or multi biomarkers are going to be how you manage, you know, yourself much better.But you know, tell me how January started. What is the thing that excited you about what you saw and what attracted you to this role? Noosheen Hashemi: Yes, absolutely. So January's origin story started with me deciding in 2016 to start my own company, essentially, after many years of running a family office, investing in, serving on boards of companies and nonprofits.  I had early success at Oracle  where I rose basically from the bottom of the organization in 1985 to vice-president by age 27. Along [with] Mark Benioff, who at the time was 26. It was quite the time, taking the company from $25 million to $3 billion in revenue. So  you know  a really, really amazing tenure there. In 2016, I started this massive research in, into theses that were getting a lot of attention, you know, big trends over the next decade. And most importantly, what I really knew. You know, the classic kind of [inaudible].  I happened to attend a conference, a White House Stanford University conference on societal benefits of AI and how to integrate sort of ever-changing AI into everyday life and into the real world. It was a healthcare panel that took my breath away. So Faith A. Lee who had organized the conference with Russ Goldman. They suggested that interested parties run off to this machine learning and healthcare conference in LA two weeks. I immediately booked my ticket. And there I met Larry Smarr. I don't know if you've come across him or not, but he was the first quantified self, maniacal quantified self person I had come across. And he had diagnosed his own Crohn's disease way before symptoms had manifested. And so, and then the common theme of this conference, between all of these presentations was that machine learning could essentially fill in for missing variables in research, not just going forward, but going backwards. So I was just hooked and I never looked back.But it was a hard problem. My own husband had been investing in healthcare and warned of like an opaque sector. He was like, “Honey, this is heavily regulated incentives are aligned with acute disease, not with chronic disease, not to mention even anything or prevention. It's just not a market economy.” And he knew how interested I am in market economies. My first love before medicine was economics. So that's a whole different podcast. So he warned that I'd be sort of fighting this uphill battle, but I was not discouraged. I basically kept on researching.I came across the MIT economist Andrew Lo. I don't know if you've come across him, but you should definitely talk to him. He's brilliant. His work showed that so little research had been done compared to what we really need to do in terms of medical research. And he comes up with ways of funding, medical research, he has a lot of innovative ways that we could really change  the whole model of medical and scientific research, but it kind of became obvious to me that the answer was that we needed to get everyone involved in research.So just, just putting things in perspective. After Nixon declared a war on cancer 50 years ago, we now have some therapeutics and some solutions to cancer. We have really nothing for neurological diseases. We're spending over $300 billion just on symptoms of Alzheimer's— don't talk about even the cure or anything like that. We have nothing for aging, which is the ultimate killer. So it was, to me, the answer was obvious, which was, we have to get everyone contributing to research. Everyone should be looking at themselves. And then with the data, we can also learn across populations. And so deep phenotyping of the population sort of in a multi-omic way was the answer.And that's what led me to Mike Snyder. I actually looked for multi-omics. I went to Stanford medical school and I met with the CEO. He said, what are you interested in? I said I'm interested in multi-omics. He said, you have to talk to Mike Snyder. And so  basically what Larry Smarr had done at the [San Diego Supercomputer Center] was to measure everything by himself. But Mike had essentially extended this kind of research to others, not just to himself. So not only sort of diagnosed himself with diabetes before the doctors, but he'd also run the Human Microbiome Project, the IPOP study, innumerable other research using metabolomics, proteomics, transcriptomics, wearables, and so on.So he had spent a lifetime studying how people went from healthy to disease essentially. And he had taken a whole person approach, which is what I was interested in.  And so in his role as chairman of genetics at Stanford and head of precision medicine at Stanford, he was kind of already living in the future. And that's kind of where I thought, you know, all of us needed to go. So our first meeting was supposed to take 45 minutes. It took 90 minutes. And in our second meeting, we agreed to join forces. It was like, it was instant.  It was just instant chemistry. Like the universe just brought us together.And then all of a sudden sort of everything fell into place for me. Looking back at my life, I been getting ready for this actually all along. Caring for my dad who had been diagnosed with cancer too late to actually give him  a surviving chance. My mom  had been misdiagnosed with asthma when she had heart failure. So I had to leave my family, you know, everyone get together and really intervene. Really changed her, her lifestyle in order to save her life. She is thankfully now 91 years old and living fine, but it has absolutely no salt in her life and a completely different, different life. My own health, my own health journey sitting in front of a computer for three decades, more than three decades, as we know that now they call it called sitting, you know, Harry Glorikian: Right, the new smoking. Noosheen Hashemi: The new smoking. My experience running a couple of hardware companies, my love of food, and my skills of kind of scaling companies. You know, all of this came together. I just basically became obsessed with prevention and I felt that, you know, food could play an outsized role.So wearables, you know, give you signals from the body continuously, which is incredible.  But you also need to understand what people are eating and, you know, we can talk about that a little bit later, but we can basically now imagine predicting chronic conditions, much like Larry and Mike had. And then, you know, postponing and potentially preventing them. And if they've already started, prevent them. Harry Glorikian: Yeah, I was lucky enough to be there and help when Evidation Health was getting off the ground and, you know, once we started to see the data coming in, I remember looking at the data. Is that real, like, is that actually happening? And I was like, the first thing I was thinking of was like, how do we design a clinical trial? Like if you're going to actually say that's happening, that trial is not going to be trivial to set up, to make that claim, but you could see it in the data.And, you know  I actually think some of the shifts that you're talking about, if it wasn't for things like the Affordable Care Act, if it wasn't for putting EMRs in place, if it wasn't for some of these shifts that have happened, you and I would still be, you know, battling this system that pays you no matter what. Right? And I think now is technology is a way that that can empower the average person to manage their own health. I'm not going to say optimally, but boy, a hell of a lot better than no information. I mean, at least some information can maybe give you an early warning light of something that you might be able to intervene in.And I don't know anybody that likes being sick. I mean, I don't do well when this thing starts to age a little bit and not function the way that I want it to. So I've tried to try and keep it in as good of a running condition as I can. So it lasts as long as possible. I mean, I'm one of those people that would listen if I just drop dead at 95, like just boom gone. I would be so happy. Right. As opposed to this sort of chronic  dynamic. [musical transition]Harry Glorikian: I want to pause the conversation for a minute to make a quick request. If you're a fan of MoneyBall Medicine, you know that we've published dozens of interviews with leading scientists and entrepreneurs exploring the boundaries of data-driven healthcare and research. And you can listen to all of those episodes for free at Apple Podcasts, or at my website glorikian.com, or wherever you get your podcasts.There's one small thing you can do in return, and that's to leave a rating and a review of the show on Apple Podcasts. It's one of the best ways to help other listeners find and follow the show.If you've never posted a review or a rating, it's easy. All you have to do is open the Apple Podcasts app on your smartphone, search for MoneyBall Medicine, and scroll down to the Ratings & Reviews section. Tap the stars to rate the show, and then tap the link that says Write a Review to leave your comments. It'll only take a minute, but it'll help us out immensely. Thank you! And now back to the show.[musical transition]So you mentioned AI, you mentioned machine learning. Where do machine learning and other forms of AI fit into January's service and you know, what do you do on consumer data? What kind of predictions can you make that wouldn't otherwise be possible?Noosheen Hashemi: Okay. I can first talk about exactly that. What did we do that hadn't been done before. What is really unique? What are we filling? So essentially in one word, it is prediction. You said it. So  as you know, there've been, there have been glycemic prediction models for type 1 diabetes, but type 1, as, you know, is a serious condition, which, you know, precision really matters for type one. It's life and death.But there hasn't been much done with type 2 diabetes. And so we set out to do predictions, for type 2 diabetes. And  the type 1 diabetes models are pretty simple. They basically are an insulin-carb calculus, essentially. But as we dug in, we realized that  you know, carbs are not all the same and that there are so many other factors besides carbs that affect glycemic response, including things like fiber fat and protein, water, and foods. We wanted to understand glycemic index and glycemic load of foods. So our major  machine learning  research projects, we basically did research for two and a half years before we sold anything.  One of the first things that we did was to try to understand the foods themselves. So we essentially built the largest database. Essentially we licensed all the, these curated food databases, and then we labeled the foods that didn't have food labels, because right now the only food labeling you really have is like grocery foods and chain restaurants.So we labeled foods and then, recognizing that glycemic response was better associated with glycemic index than carbs alone, we set out to create glycemic index and glycemic load for all these foods. Then we ran a clinical trial and associated people's glycemic response to the glycemic load of foods they were eating. And then we turned that into a prediction. So, the prediction model.  Why is it so cool? Well, why should you use your body to figure out how many glasses of wine is going to spike you? Why not have the AI tell you that? Why not do that in silico?  It's this weekend, you want to cook for your wife. You want to get her the right fried chicken recipe. Well, check those out in January, check out those recipes in January. If you know what the glycemic response of, of each one of those recipes could be, it really helps you compare foods. For kind of recipes you can comparefood items in your local cafe. You want to figure out what to eat. You don't have to put them through your body to figure out how you're going to respond, put them through the AI to figure out how you're going to respond.And then in terms of, you know, how we're different. I mean, we essentially live in the future. We, we don't  we don't live in blood pricks and strips and blood glucose meters. We kind of live in the CGM, HRM (heart rate monitor) precision foodworld. We've turned food into actionable health data, which is a necessary ingredient you need if you want to understand people's glycemic response. And if you want to be able to predict it, and that is our huge innovation that nobody has. And we have quite a bit of IP around it. There are a number of things that we're using. We're using meta-learning.  We're using  neural networks. I don't know how much I should say about what we're using. Yeah. We have one paper that we've put out, which is really, really, really simple.  But we, we always talk about, what kind of papers we want to put out and how much we should put out and how much should we not put out, but essentially you can look at the people that advise the company and you can see that, you know, we have a lot of expertise around  essentially… Harry Glorikian: But Noosheen, when you're doing this right, you need to, at some point, I think you need a baseline on say me for a certain period of time before the algorithm can then respond appropriately to that. And then doesn't that potentially change over time, time you mentioned the yogurt, the meusli, right.  And how that affects. So it's constantly gotta be in a feedback learning loop.Noosheen Hashemi: Yes. Yes. And the beauty of January is that essentially you don't have to wear a CGM 365 days out of the year. We think that with AI, we allow you to wear a CGM intermittently. So maybe you want to wear it every quarter  to update our models  just to see how things are going, but you don't need to wear it all the time. You can wear it for a period of training and then basically run your simulations in silico rather than through your body. Let the AI do the work. So you definitely should wear it intermittently so we can update our, our models because people do age. People do have inflection points in their health. They get pregnant, they travel, a lot of things change, but we don't think it's necessary for healthy people to wear CGMs all year long necessarily. Harry Glorikian: So now we're talking about consumer behavior, right, for a, for a tech product like this. And if, you know, if you look at some of the data that I've read in some of these papers, you know, the potential market is significant. It's, you know, it's quite large. I mean, if I just said, you know, 15% of the people have pre-diabetic levels of glucose after eating, that would translate to like 50 million people in the United States alone. But the service depends on the CGM, the app, the external heart monitor. It's, you know, users have to be diligent about monitoring and logging food intake and activities during the introductory month. So for a quantified self junkie, I get it. They're all over this.  What's the plan for getting everybody else on to this? Noosheen Hashemi: Well, I think it's all about the user experience. And I think we have a, we have a long way to go as an industry and for us as well.As a company we have, what we imagine to be the user experience is nowhere near where we are today.I'm old enough to remember world before Starbucks. So you would see ads on TV for MJB coffee, which is something you made at home. You know, I don't know if you remember that but Starbucks created a new experience, really a place between home and work where you would stop by for coffee.And so the outrage around the, you know, $3, $4 latteat the time, do you remember that?Well, Starbucks continue to improve the experience. They added wi-fi, they had ethical coffee, they had kind of a diverse employee population. People's initial wonder and worry gave way to this, you know, gigantic global brand. And I think all of that is because of the experience that people had. I think we need to make health a positive experience. We need to—we, including January—need to make health something that people….it's going to be a little clunky in the beginning, just like the old, you know, cell phones used to be. But while we're going through this process, the companies need to work on to improve the experience and people need to be patient with the clunkiness of everything  to get us to a place where these things become much, much more pleasant to use and easier to use, and essentially AI starts reading your mind about what you were eating and what you were doing. That is going to happen. You know, I've gotten so used to my Apple Watch now that I actually love it. It actually is doing a very good job training me. Just at the right time, you know, “Come on, you still have a chance. Let's go.” You know, all the things that it's doing  I'm actually liking it. It's it's enjoyable. Because it Is coaching. And I feel like the answer for mass adoption lives in experience. We need to improve the experience dramatically. Harry Glorikian: It's interesting though, because I I'm play with a lot of these different things and I noticed that depending on how they're designed, how they're put together, it nudges me to do that much more or et cetera. I don't always listen. Human beings don't always do what they're supposed to do for their better good. But  you can see how, when the app is designed in a way to nudge someone the right in through the right mechanisms. And that's the problem, right, is trying to—not the same mechanism works on everybody. So you may have to have multiple approaches that the system tries like AB testing for a website to, to get them to do that.But so, if the average person like me wants to do something like this, obviously I have to get a ‘script from my doctor, which just drives me crazy that I can't just—because I can buy a finger-prick, right, over the counter and poke myself a thousand times and then write down these numbers to see what happens. Which seems a little clunky in my opinion. But I can't buy the CGM that does it automatically. There's gotta be some medical person saying like, we're gonna make more money off this if we do this or do that, or, or it just doesn't make any sense to me.  How do you, how does January come at the expense reimbursement or the insured part of it, or is this just out of pocket for everybody? Noosheen Hashemi: Sure. So right now  government insurance, companies, and private insurance companies cover CGMs for people that are intense insulin users. So people that prick themselves four times a day. And so that's three and a half million out of 122 million people that have pre-diabetes or diabetes. So it's a very small population. And the rest is all cash paid. And it it's really out of pocket. So we have an early access price of $288. And we, you know, we include the CGM, but you can also buy CGMs only from January. You can just, if you just want a CGM, you don't want to do anything else. You're just curious. You want an introduction to this world? You can order a CGM from January for $80 if you want to do that. So if you're one of the 12 million people that are insured by Kaiser—and Kaiser doctors will not write you a prescription, you can go to your doctor and ask them, they won't write you a prescription—come to January. We will give you a CGM. You can be introduced to the program and then, you know, take, take up January from there and experience the magic of CGMs alone. I really do think they are a magical product because they they're showing you for the first time you kind of can see inside your body, which is really phenomenal. Unfortunately by themselves, they're not that effective and they're not that effective by themselves longitudinally. So if you really want to keep track of how you've been doing, what food spiked you, how you can, you know, what kind of exercise, things like that. They don't really have that additional intelligence, but they are magical, they are really magical tools. But, you know, you want an insightful experience on top of that. With the AI that can essentially synthesize this kind of data from your heart rate, monitor from your food, from your glucose monitor and sort of let you know how much to eat, what to eat, how to hack your food, how much to walk, how much, how much to fast, when to fast, how much fiber you're having, not having. That's where we come in. Harry Glorikian: I feel like at some point I'm going to need a big monitor in my house that just tells me these things as I'm walking by. But you know, it, it's interesting. I mean, we are entering the era of real wearables and apps and big data and, and, you know, but here's the question though. Soyou know, Apple just announced what's going to be the update to their iOS and, you know, pretty soon I'm going to be able to push a button and share data with my physician.  Which is funny because I go in his office and I pull up my phone and I'm like, here's my longitudinal. And here's my longitudinal. And I'm like, look, you can take the measurement because you're supposed to, but here's how it looks over the last three months as opposed to the one time when I'm here. Can January's customers export and share the data with their doctor? Noosheen Hashemi: We have a report  midstream at 14 days that you can share  with, with your doctor. But of course we intend to, you know, we have features planned that are going to make things way more easily done, much more easily in the future. We really strongly believe that people should own their own health data. We are huge advocates for people owning their own health data, because there are a lot of people hanging onto your health data and they don't want to give it to you. I'm talking about device makers and others. You're paying for the device, which comes with the data, but they don't want you to have the data. So they're like, “You can have the data and study it yourself, but you can't give that data to other people.” But that doesn't work.We are living in a multi-omics world. Single 'omics by themselves, the single side node biomarkers, you know, “Harry, you just manage your cholesterol. Noosheen, you can't keep two things in your head. Why don't you just manage your A1C? And Mike, you should watch your blood pressure.” That just doesn't work. There are many, many markers that you've just, as you just said, that we need to keep in our heads. We can't keep them in our heads, but that's where AI comes in. We need to feed them into something and people must have the right to own their data and share their data with whoever they want. If it's their coach, it's their doctor, it's their wife or spouse or significant other, their dog. They should be able to share the data that they own.As long as they provision it properly to whoever they want to give it to because you know, someone doesn't want their employer to know X, Y, and Z. Somebody else wants their coach to know that is people's rights. And coming from kind of a libertarian point of view, I really think people, you know, people should own their own data and they should be able to mix it with other data  for synthesis, if they want to. Harry Glorikian: Yeah, it's interesting. I mean, I totally believe in that. I always, I also understand that people may not understand the implications of sharing sometimes.  And that's not clear, but I do believe that the next iteration of where we're going to see this technology go is multifactorial software programs that can take a number of different inputs to give a much more holistic view of what's going on with me, so I can manage myself better share that information. My biggest worry is most physicians I know are—it's not totally like, it's not their fault, right….Noosheen Hashemi: They're so busy, so they're spending 15 minutes a year with you. And during that 15 minutes, you know, they're taking a point in time, you know, to see a snapshot of your health. And your health is way more complicated than that. We're talking about reverse engineering, 5 billion, years of evolution. And you know, they're going to get, see if such an infinite small part of that. We need to be way more self-aware.Harry Glorikian: Well, it's funny because I do have, some of my physician friends will be like, you want me to understand that genomic marker that whatever, like, I can't, I can't get my patient to manage their insulin level!Noosheen Hashemi: I have a lot of empathy for that. They just don't have the time.  I completely fully understand. Which is why I think we should carry more of the, we should have more agency over our health and we should carry the burden a little bit more.Harry Glorikian: So what is wild success for January? Noosheen Hashemi: Well, we want to keep on this path of developing our multi-omic platform. We want to essentially  help people understand themselves deeply and figure out how to dial their lifestyles and sort of tweak and tune their health. This is non-trivial obviously because there's not enough research in food science or enough research on prevention. You know, out of the $3.8 trillion that we spend on healthcare, 2.9% goes to prevention and 10% goes to acute care end of life care. Just think about that. More than three times as much goes to end of life acute care than goes to prevention. And I'm talking about healthcare costs, I'm not talking about research costs in terms of what NIH and USAID and all of those people spend. So there's not enough research that's happening. You know, people's health data is not organized today. I'm sure there are companies who are trying to organize the world's data. You know, the company that tries to organize the world's data is trying to organize your health data. So I think that's pretty smart.  I think today it's still very opaque and it lives in silos, but I think in the future is going to be mixed.  I think today people just aren't fully empowered yet, you know, with the knowledge and with the agency and with the tools they need to really manage their health.Wild success for us means that people, that we're part of this revolution of consumerized healthcare. We're part of the food-as-medicine revolution, the precision nutrition revolution. So we see ourselves coming up with tools that can essentially get amazing experiences in the hands of millions of people.If you can think about a company like Livongo going public with 192,000 patients. Or if you think about everyone that's playing in the metabolic health today, if you put 12 or 13 companies together, maybe they have a million users, or maybe a million and a half users. Where is that compared to 122 million people that have pre-diabetes diabetes and another a hundred million people that are optimizers? They're either wearing a wearable, they belong to a gym, they're on a diet. You have the entire population as your market. And we have very little that has really made a major foray into health. So wild success means having a product that becomes mainstream. Harry Glorikian: So I think what you're saying is January is moving beyond just CGMs and metabolic syndrome, right?Noosheen Hashemi: Absolutely. Yeah, we, we imagine ourselves, we have built an expandable platform. Our goal is to keep doing deep phenotyping. So we will add 'omics  you will see us adding 'omics beyond what we have today. You will see us  get to other cardio-metabolic disease, you know, cardiometabolic disease, essentially going beyond metabolic disease to the rest ofmetabolic syndrome. You'll see us be ahardware-agnostic company. We want to essentially let people wear whatever they want.  Whatever works for them and, and still try to bring that data, synthesize it and make sense of it and feed it back to them so they can take action. Harry Glorikian: Excellent. Well, that's, that's a great way to end the program with. We have so much more to see from the company and what it's going to be able to do with the data and, and, and help  you know, people live a healthier life. Or like I said, with me I'm constantly trying to measure what's going on. It's just distilling it to make it easily consumable to do what I need to do rather than have me learn statistics so that I can figure it out. Noosheen Hashemi: We have to get, all of us need to get better than that. I remember when I first put on my Oura ring, you know, there's, you know, most people  first when they wear their Fitbits, you know, first it was like, how much did I sleep? And then they kind of learned about REM and sort of deep sleep and then slowly. And then Oura came and then it was like, oh, and Whoop had already had heart rate variability, but then, you know, Oura came in with their other markers, you know, restfulness. And efficiency, sleep efficiency and timing, et cetera. And so people are slowly wrapping their heads around this. It takes a little whil. And yes, January gives you a lot of levers. You know, there's fasting, there's fiber, there's calorie management. There's you know, the spikers. There is the activity counterfactuals—I ate this, but had I eaten this other thing, this would have been my glycemic response. Or had I walked X number of minutes after that, this would have been my glycemic response. At the beginning it's a lot, but that's where it goes back to the experience. We must make the experience enjoyable and better, and we must, companies like us should strive to make the experience enjoyable, make them fantastic consumer experiences like Apple products. But remember Apple's 45 years old and we're just getting going with this, But [Apple is] a great role model. Harry Glorikian: Wellyou know, my doctor may not like it, but I may have to get one of these. He's listening to this podcast. I know that he will, because he always comments on them. Noosheen Hashemi: We're definitely doing that. And you know what? You can have Mike Snyder, you can chat with Mike  about your numbers after. That would be a lot of fun.Harry Glorikian: Excellent. Oh, I look forward to it. So thank you so much for participating. Noosheen Hashemi: Thank you, Harry. It was pleasure.Harry Glorikian: That's it for this week's show. You can find past episodes of MoneyBall Medicine at my website, glorikian.com, under the tab “Podcast.” And you can follow me on Twitter at hglorikian.  Thanks for listening, and we'll be back soon with our next interview.

Ni Hao everyone! After months of waiting, we're thrilled to finally introduce Helen Yang in today's episode of In the Suite! Helen Yang is an inspirational and giving person. She's a CFA by designation, a FinTech veteran for 20 years, a co-winner of the Harry Markowitz Award, an involved mother of two children, and an active board member of Asian Americans for Equal Rights. She started her career at Thomson Reuters and Charles River Development. Helen was dissatisfied with the ‘cookie-cutter' service she was receiving from financial advisors. She realized that financial advisers didn't have the right tools to deliver the kind of personalized insights and services she was looking for. With the mission to fill this gap, Helen founded Andes Wealth Technologies.For the second year in a row, Andes Wealth Technologies was selected as a finalist in the 2021 WealthManagement.com Industry Award in the Risk Tolerance and Client Profiling category, and for the first time, a finalist in the Portfolio Analytics category. Helen is inspired by acts of kindness around her and volunteers extensively to give back to society. She serves on the board of Lexington Lions Club and Asian Americans for Equal Rights. In this episode, you will learn about her awe-inspiring growth journey and how she balances her personal and professional life. We also cover powerful and educational topics, including the identity and challenges experienced by Asian Americans and what we can do to combat hate. Helen also shares knowledge on why risk tolerance is essential and shares relevant stories.Find the agenda of our conversation with Helen below:About Andes Wealth Technologies (02:57)How the pandemic made advisors look at risk tolerance differently (04:42)Congratulate Helen on her recent success (06:20)The importance of risk tolerance tools explained (09:20)The backstory behind the company's name (14.40)Sharing the Harry Markowitz Award 2011 with Dr. Andrew Lo (16:06)Helen's impressive educational background (21.19)Inspiration behind her career choice (22:08)Helen's advice to women who want a career in Fintech (25:30) Being an advocate for equal rights for Asian Americans (32:30)Educating herself and others on Asian American history (38:00)How can we support Asian Americans? (40:06)“Good Deeds For Good Karma” (44:22)Speaking at the Wealthies Forum (47:00)Book Recommendation by Helen (55.38) Behavioral biases during the pandemic (1:03:19)Resources-Helen Yang Twitter-Helen Yang LinkedIn-Andes Wealth Technologies LinkedIn-Tsinghua University-Andes Wealth Technologies-Andes Wealth Technologies Named Finalists for Two 2021 Industry Awards-t3 Technology Software Survey Report-The Undoing Project: A Friendship that Changed Our Lives by Michael Lewis on Amazon-Growing Tree

In this episode, Andrew Lo shares some insights from his journey with innovation. Through his work, Lo utilizes financial engineering to tackle some of the greatest challenges we are facing today such as Cancer, Climate Change, the Financial Markets, the Pandemic and its associated economic challenges and opportunities. Andrew W. Lo is the Charles E. and Susan T. Harris Professor, a Professor of Finance, and the Director of the Laboratory for Financial Engineering at the MIT Sloan School of Management.

In episode 310, we welcome our guest, Kathryn Kaminski, Chief Research Strategist at AlphaSimplex, where she’s also the co-portfolio manager for the firm’s Managed Futures Strategy.     In today’s episode, we’re talking all about trend following and managed futures. You may have heard the phrase “crisis alpha” before, and Kathryn is the person who coined that phrase.  We start with hearing what it was like for her to study at MIT under the legendary Andrew Lo. Then she explains why trend following works during a crisis and uses last year as an example.    As we wind down, Kathryn explains some misconceptions about trend following and talks about why it’s so important to have a process driven investment approach.    All this and more in episode 310 with AlphaSimplex’s Kathryn Kaminski.   ---   This episode is sponsored by Masterworks. Masterworks is opening the doors to top-tier, blue-chip art investments to everyone. Use Promo Code “MEB” to skip their 15,000 person wait list.

The family of Wan Li Zhu emigrated to America seeking new opportunities. Wan Li benefited from high-quality public education at Bronx High School of Science and went on to a perfect grade-point average at MIT. He studied under renowned quant wiz Andrew Lo and was poised for a career on Wall Street but was lured away by the prospect of hands-on responsibility for product features at Microsoft. After a prodigiously successful stint, during which he was involved in building and marketing Dynamics CRM, Microsoft’s fastest-growing product, he went to Harvard Business School. From HBS he was recruited by early-stage VC firm Fairhaven Capital.  The firm, known for its expertise in web security and digital advertising, now sees promise in various applications of artificial intelligence starting with self-driving technology. Wan Li is deeply engaged in bringing on the next generation of winning investments at Fairhaven Capital. Despite a busy professional life, Wan Li Zhu has found time to advise startups and to co-found MIT Angels in Boston. I learned a ton from my conversation with this wise, yet unassuming early-stage VC and angel.  HERE'S A LIST OF THE TOPICS COVERED: Wan Li Zhu Bio Studied with MIT Professor Andrew Lo – Used Natural Language Processing to Assess Market Sentiment Why Wan Li Zhu Went to Microsoft – Three Years at MS – Shipped Three Versions of the Product Wan Li Zhu Connects with Fairhaven Capital through HBS Resume Book Fairhaven Capital Is Thesis-driven – Attentive to Market Trends that Could Create Large Opportunities How the Fairhaven Capital Portfolio Is Doing What Wan Li Zhu Looks for in a Startup Investment Experienced Founders Can Actually Time Markets TVision Came Via MIT Angels – Measuring Engagement of TV Viewers AirFox – Enabling Wireless Carriers to Offer More Affordable Data Plans MIT Angels Company PathAI’s Deep Learning System Is Better at Detecting Tumor Cells than Human Pathologists Latch – Enterprise-grade Keyless Access System for Apartment Buildings The Investment Wan Li Zhu Regrets Not Making Wise VC Wan Li Zhu Continues to Be Very Bullish on AI

The journey of music is special, the stories that can be told without saying a word. Music is special that way. Tune in as Darko jumps in and out of different worlds brought to you by the various beatmakers. Yes as always it's going to be a roller coaster. This episode is dedicated to my Uncle and to DMX, may they rest in paradise.  Andrew Lo - 04:09 , haruka nakamura - 12:22 , siyasha - 17:09 , Paul Alamwala - 20:35 , Smyth - 28:48 , Mani D - 35:54 , Xcephasx - 43:21 , Skinny Local - 48:17 , Andrew Lo - 52:00 --- Support this podcast: https://anchor.fm/chillroseradio/support

Andrew Lo discusses his various fields of expertise and research with Anastasia Diakaki. Professor Lo's most recent work in healthcare finance, more timely than ever, serves as the starting point for a conversation covering the COVID-19 pandemic, machine learning applications in biopharma, artificial intelligence and the interpretability of machine learning results in the finance industry. When it comes to preparing for the future, we ask Andrew Lo: "Do we all need to become computer scientists?"   Topics discussed: 1:30 - Professor Lo’s philosophy and approach in health care finance sector 2:43 - Applications: what are the implications of these tools? 4:51 - Lo’s view on the pandemic and the acceleration of some processes 6:41 - Machine learning applications and health care 8:58 - Successes and improvements to algorithms in practice 10:57 - Computer science – the new life skill to have 13:24 - Machine learning algorithm used in the investment process 15:25 - Are your current students more comfortable with science skills? 16:55 - AI and human intelligence   Related articles and links:   CFA Institute members and charterholders Professional Learning Tracker https://cpd.cfainstitute.org/ ____   ► Subscribe to the Take 15 channel here https://www.youtube.com/channel/UCA3HUMuK4FSp_CvQH_2Ji7g   ►Email us for comments, questions, or appearance requests:Take15podcast@cfainstitute.org   ____   Find Andrew here:   LinkedIn: https://www.linkedin.com/in/andrewwlo/   Twitter: https://twitter.com/AndrewWLo?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5Eauthor   Find Anastasia here:   LinkedIn: https://www.linkedin.com/in/anastasiadiakaki/?originalSubdomain=uk   Twitter: https://twitter.com/anastdiakaki -- Find Lauren here:   LinkedIn: https://www.linkedin.com/in/lauren-foster/   Twitter: https://twitter.com/laurenfosternyc -- Listen to us on:   Libsyn: http://take15audio.cfainstitute.libsynpro.com/   Spotify: https://open.spotify.com/show/4z0UYTI6B1pfSk5UrpCqFk?si=VzxBsKUWTme_D9PSk25r3g   Apple Podcast: https://podcasts.apple.com/us/podcast/take-15-podcast/id268942353

In today's episode we talk about how to be a rational investor. We have relied on excerpts from “Adaptive Markets: Financial Evolution at the Speed of Thought” by Andrew Lo and Nate Silver's “The Signal and the Noise: Why So Many Predictions Fail-but Some Don't”, to put this story together

Om makt och pengar bakom den globala vaccinkampen och varför även vinnarna riskerar att förlora när mutationerna sprider sig över världen. Medverkande: HannahKuchler, journalist som bevakar läkemedelsindustrin förFinancialTimes, Noubar Afeyan, styrelseordförande och en av grundarna av Moderna, Andrew Lo,professor i finansiell ekonomi på MIT, Andrea Taylor, forskare vid Duke global health institute, Susanna Zeko, generalsekreterare för Internationella handelskammaren ICC i Stockholm, Sevcan Yesiltas, assisterande professor vid Koc universitet i Istanbul, Pedro Villardi, aktivist som doktorerat på patenträttigheter, José Gomes Temporão, fd hälsominister i Brasilien, Thomas Cueni, ordförande läkemedelsbolagens internationella samarbetsorganisation IFPMA, Mustaqeem de Gama, Sydafrikas chefsförhandlare i WTO, Zakaria Gansané, epidemiolog i Burkina Faso, Charlotta Zacharias, vaccinläkare mfl Programledare: Ivar Ekman ivar.ekman@sr.se Producent: Lotten Collin lotten.collin@sr.se Reportrar: Ulrika Bergqvist och Anja Sahlberg Tekniker: Fabian Begnert

From Hong Kong, Andrew Lo, Chairman and CEO of EFT Solutions, shares insight on what is going on with payments and commerce and the future of fin-tech in China.

In this conversation I talk to Luke Constable about the complicated tapestry of finance, funding projects, incentives, organizational and legal structures, social technologies, and more. Luke is the founder of the hedge fund Lembas Capital and publishes a widely-read newsletter full of fascinating deep dives. He’s also trained as a lawyer and historian so he looks at the world with a fairly unique set of lenses. Disclaimer: nothing Luke says is an offer to buy or sell a security or to make an investment Links Luke on Twitter Lembas Capital Theory of Investment Value (John Burr Williams) 1,000 True Fans (Kevin Kelly) Quantum Country Patreon Lembas Capital’s Open Questions The Empire of Value (André Orléan) Who Gets What and Why (Alvin Roth) The Mystery of Capital (Hernando de Soto) I, Pencil (Leonard Read) The Crime of Reason (Robert Laughlin) Andrew Lo’s papers Transcript 0:01:05 BR: So if technology creates a lot of wealth, why does it feel like most people in finance are hesitant to invest in technology?   0:01:19 Luke Constable: So that's an interesting place to start. I think you have to understand, no one invests in technology. If you think about investors, investors invest in businesses that use technology, and so that's probably the first frame I would use. Investors aren't hesitant to invest in technology, investors never invest in technology. What investors do is they invest in these products that are going to generate cash flow streams, and so that's sort of the first thing. And then the second thing is, a lot of the technologies that you and I think about, they seem obvious at a macro scale, where you take a high level view and you say, "Well, it would be so much better if we had a blank sheet of paper," and I said, "We should do X."   0:02:10 LC: For instance, you could make an argument about housing technology in San Francisco, and you could say, “All of these houses built in SF, they're old Victorians, they don't really have washing machines and laundry machines, you could probably change the structural engineering, probably build them higher”. And if you look at them and said, "Oh, I have a better prefab housing technology," or "I have a better way to do it," you'd miss the point, which is just because you've invented the physics, and this is the other thing, you actually have to sell it into a market. You have to work within the market, and so that's usually where I see a lot of the interesting technical products fall down.   0:02:53 BR: So the thing that I want to poke at in the assertion that people invest in businesses is that people invest in things that are not businesses as well, people invest in gold, in currencies and other, I guess, assets would be the high level thing, and so I guess the question is why isn't technology itself an asset, and there's probably a very obvious answer to this, I just...   0:03:25 LC: Sure, so let's take a step back and talk about the various asset classes, there's sort of a couple of ways to break them down.   0:03:32 BR: Okay.   0:03:33 LC: One way people do this is they'll say there are real assets, these are things like real estate, some people put commodities in there, and then there are sort of these yield assets, these are debt that is putting out a cash flow stream, and then you have equities, and there's some argument that cryptocurrency is sort of its own asset class, and then currencies might be their own asset class too. And what you'll quickly find is these things kind of blend together. A lot of them are different ways of financing sort of the same project. And then you have the ones that are just traded for their own sake. So there's sort of two questions you're asking, the first is, why isn't "technology" the same as like gold or silver or real estate, for instance? And so there's a use value to all of those commodities, and that's why they have value, and that actually is a cash flow stream, we actually do use gold, we do use silver, and that's how that works.   0:04:43 LC: But if you think about what's valuable, there's sort of something that's value... And I should have started with this. When you think about what value is, there's value in exchange and then there's value in use. So the value in exchange ones, these are often, you could argue, cryptocurrency or a lot of currencies, gold is actually usually thought of as a medium of exchange, that actually is valuable for cash flow purposes just probably not in the ways that you think. So what happens with these currencies and these stores of value is they sort of become Schelling points where I just know there are enough people transacting in that thing that I can find the liquidity, I can actually go convert to cash, and I can go basically get that cash when I need it. That actually is a cash flow need. It's just not often thought of that way.   0:05:40 LC: Now, liquidity is really valuable because you might be invested in the best business of all time, and it might have a very, very, very high net present value and be doing a lot of good for the world. But if you take a step back and say, "Wait a second, I have to pay off student loans," or "I have to pay off my mortgage," or "I just want some cash to go on vacation" or whatever you want to do with it, you look at this and say, "Gosh, I do need some liquidity," and that's what those other sort of trading assets are for.   0:06:10 BR: So basically, technology contributes to the use value of an equity asset, is that the right way to think about it?   0:06:22 LC: I don't think of technology that separate from... It's sort of so baked into the environment that it's just difficult to disentangle. Technology, lazily put, is just ways of doing things hopefully more efficiently than we're already doing them. And so if you think about why certain assets become tradable, either they're creating these cash flow streams, or there is some value in exchange. I mean, the way that I often frame investing for the people who I invest for is there's sort of two sets of flows that determine an asset's price. There is underlying asset's cash flows and then there are the capital flows of all the investors. So you have sellers for some reason, maybe they have liquidity needs, maybe they can't hold an asset for a regulatory reason or a legal reason, and then you have buyers who come in, because they're interested in that asset, and it could be because they think it's an interesting thing to invest in, it could be because the regulators told them that they have to buy it, it could be... You laugh, but this is actually...   0:07:32 BR: What sort of things do regulators mandate that people buy?   0:07:37 LC: Sure, so if you go look at banks and sovereign debt, well, actually banks and all debt. So you have the bank regulators set risk weightings on various types of debt, which is sort of a nice way of saying, there are all of these different cash flow streams, and the regulators are saying to you that certain cash flow streams are riskier or less risky. And shockingly, they often argue that their sovereign debt is less risky than some other cash flow streams.   0:08:13 BR: I'm shocked.   0:08:14 LC: In practice, that may or may not be true. It's a weird thing to think about, but, in some cases, a multi-national corporation might actually be a better credit than a country. But that's not how these things work, and so what happens is a bank regulator will sometimes go to a bank and say, "The risk weighting on the sovereign debt is far lower than the risk weighting on this corporate debt,” which effectively is pushing the bank to go buy a certain type of debt, which then goes and funds all of those projects. So then coming back to all of this, if you think about investing in sort of these two sets of flows, like that underlying asset's cash flows and then the capital flows of all the investors, you basically, in practical terms, want to think about markets in terms of what's driving someone's action.   0:09:05 LC: And when you think about that, that's when market prices start to make sense. They won't make sense to you if you think that you're just going to sit down and solve an analytical equation where you just sort of put in a few inputs, you make a few estimates and then the price gets spit out. It's much more of a socially constructed thing.   0:09:25 BR: And going back to your point about liquidity, it feels like there's this... I don't know how to describe it, like sort of a weird effect where it feels like there's a consensus that investing in... I won't say technology, I'll say investing in a business that is proposing to build a technology with a very long-term time scale, there's consensus that that will eventually create something... Will eventually create a lot of value, but then at the same time, because of these liquidity constraints, very few people are doing that, and that's the argument for why people are not making those investments, but it seems like that would be a point where you could arbitrage. It seems like there should be some people who are willing to not get cash flow for a couple of decades, and they would be able to reap the rewards of making these sorts of investments, but you don't see that, so I assume that those people are smarter than I am. And so the question is, why don't you see people doing that?   0:10:50 LC: So you actually do see people doing this literally all the time, but it's not for the sexy technology concepts that you are thinking of. So go look into the public markets right now. You'll see a handful of software businesses that are trading at very high multiples to sales. So the idea is that you sort of have this trade-off: you could get free cash flow after taxes right now, or effectively more free cash flow down the line from some company that's growing quickly, and so what you do is you pay some price based on that free cash flow multiple. What happens when the free cash flow is really, really far down the line, we don't even use the free cash flow number, we actually just use the sales number. And sales is obviously much higher than just free cash flow, 'cause free cash flow is after all of your expenses and taxes. So when you go look, and you see some company that's trading at 15 or 20 or 25 times sales, the stock market is betting on that business being around and generating free cash flow over a 25 or 30-year period. That's the only way that math works. In practice, the reason the stock gets priced that way has something to do with those cash flows and also a lot to do with the capital flow landscape, but that is what's happening.   0:12:15 LC: These companies are getting funded on a 30-year time scale, and so the right question shouldn't be, "Why aren't good projects getting funded?" They actually are. The right question is, "Why aren't other good projects getting funded?" And so I think it comes down to... I think it comes down to what is legible to institutional finance, and so you might look out into the world and say, "There are trillions of dollars of capital... " I mean, there's just oceans of money out there, and it seems like someone could raise billions of dollar to go trade a building with someone else or something else that seems like it isn't actually moving the world forward and this sort of simplistic take. But why can't we take that billion dollars and put it towards some technology, something that might be obvious in your opinion toward moving the world forward?   0:13:15 LC: So the first thing is you have to understand what matters is, in practice, even though it looks like there are trillions of dollars of capital out there, risk-adjusted or uncertainty-adjusted, there's actually very little capital available. And the right way to think about it is to say, what type of product are the capital allocators buying? And so this isn't, again, a place where we have an analytical equation and you just pop your numbers into the equation and you say, "Well, the return to society would be X percent higher if we invested in this type of technology that will have a payoff in 25 years." The right way to look at it is to have empathy with the person who is in this capital allocator's seat, in this investor's seat...   0:14:08 BR: I.e you.   0:14:08 LC: Well, me or anyone else. But again, I'm not trying to paint myself upfront, there's the intellectual side of capital allocation, and then there's the reality that a lot of people are using an element of gambling in this. But it's to understand what they're buying. And so the reason people are comfortable investing in that real estate or investing in an enterprise software company is someone has come up with a set of metrics that has convinced the market that those cash flow streams are durable, that they will exist and be predictable 20 or 30 years out. And so what you've done is you've created this yield product, and what you've really done is you've created a sense of certainty. And I think what people don't like is uncertainty, they really want to essentially have something that they don't have to do too much intellectual work to understand and that they feel like they can trust. And so the problem is actually sort of one of search costs.   0:15:20 BR: A really dumb question is, What does it mean for something to be risk or uncertainty adjusted? Because you said that there's trillions of dollars out there, but there's actually not that much when they're risk or uncertainty adjusted, and is that basically just say that capital allocators don't have the incentive to spend most of that money on anything that they perceive to be risky or uncertain?   0:15:50 LC: Not exactly.   0:15:51 BR: Okay.   0:15:52 LC: It's two things. So first, in terms of how most people think about risk, so the way that you might think about this before you start really looking at it is you'd think, Well, we're just trying to sort of predict the future, the future is relatively predictable, and we can make some educated guesses about probabilistically what is going to happen, and then we can sort of model out those payoffs, those defaults, and sort of go from there. And so sort of the canonical text in finance for equity evaluation is called The Theory of Investment Value, and it's written by a guy named John Burr Williams. I can send you links after this. It's written by a guy named John Burr Williams after the Great Depression, and he was basically trying to sort of scientifically estimate the value of all free cash flows. You may have heard of this concept of discounted free cash flows?   0:16:48 BR: Yeah.   0:16:48 LC: He's arguably the person who invented it or at least codified it. In practice, though, you quickly find it is unbelievably difficult to figure out and to actually estimate the cash flows of something, even four, five or six years out. The world just changes really quickly, competitive positions tend to change really quickly, and so you actually could come up with this range of outcomes, but they become somewhat uncertain. So you take that as sort of the investing reality, and now let's look at sort of the funding reality. A lot of the people who fund investment funds or who are making investments, they have cash flow needs. They have sort of real cash flow needs, and then they have sort of intellectually forced cash flow needs. The real cash flow needs are, look, we have to fund our endowment, we pay X percent out per year so that the college can function, so that the hospital can function.   0:17:53 LC: And then the intellectual cash flow needs are, look, here are the risk models that we use, and when we see the prices of our investments fall 8%, we consider that as fundamental information that our investments aren't performing well, and so we need to sell out. And so they actually don't just need cash flow to look good, they need the pricing information in the market to look good. So we're talking about arbitrages. This is probably one of the biggest arbitrages that exists in the market, but it's unbelievably difficult to capture. So let me give you an example, imagine that you had a row of 10 houses in a neighborhood and they were all... Let's just say for these purposes, valued at $100. So let's say one of the neighbors, they are in a rush and they need to sell their house because they got a good job offer somewhere else, so she sells her house for $97 because she'll just get whatever she can get. And then another neighbor gets a similar job offer, and she sells her house for $95, and suddenly some other neighbor along the street looks around and says, "Oh no, prices are falling on our houses, everything else is getting sold off, we need to sell." And so they might sell just because they're scared, because they think there's sort of fundamental information in those transactions, in saying, "Okay, the market price has fallen."   0:19:22 LC: So you've seen the marked prices fall from a $100 down to $95. The problem is the market shows the prices of transactions, they don't necessarily tell you the fundamental value behind those transactions. So as a result, you being a portfolio manager, say you're invested in houses, you might have a view and say, I think that those houses that sold off, those were forced sellers. That doesn't mean that the price of the assets have actually fallen, these prices will come back up. Someone else might say, "No, no, that's pretty arrogant of you. The market has spoken and job opportunities have changed and people are going to leave the neighborhood." Now, it's really difficult to capture that sort of arbitrage, and arbitrage isn't even the right word, but capture that valuation spread, because it actually comes effectively down to who is right, and that ends up being a grounded matter of opinion, but effectively a matter of opinion.   0:20:32 LC: You can do a lot of diligence, and then you can maybe figure out if you're generally more right or generally more wrong. Ideally, if you get really, really good at sourcing information on the asset cost that you're investing in, and then you go around looking for these situations where the market has sold them off, but you recognize that they're sort of incorrect in doing it. But for the big portfolio managers, again, there's an information search cost. Every single time one of their fund managers underperforms, fund manager is of course, going to come back and say, "No, no, it's temporary. We're right, the strategy will come back. Don't pull your money."   0:21:12 LC: And so the difficult thing for the allocators to funds is they sort of have to diligence the fund managers who are then diligencing the investments. And so you can see that as you sort of go down this line of information being passed from person to person, the search costs just rise. Whatt it really comes down to is basically trust, where the investor is investing in a company or in some operator, and then the allocator is investing with the investment fund. And all along those links in the chain, it's so expensive from an information perspective to figure out who's being honest and who isn't. That trust is actually the fastest way to figure out what is a good investment and what isn't.   0:22:06 BR: Yeah. Correct me if I'm wrong, but then I sort of extrapolate that to the thought that it's actually very hard to build up trust in someone who's proposing to make, say, a 25-year bet, because you would need 25 years to build that trust, right?   0:22:31 LC: Sure, and this is actually the problem. And so if you look at it, most fund cycles for the investment funds themselves, they typically have about a three-year window to prove themselves. So if they can't show marked prices rising within two to three years, or they can't show cash flows coming out in those two to three years, it's in practice really difficult for that fund manager to go raise more money from an allocator. The best allocators, they really get it. But in practice, most people are sort of looking at each other trying to understand what we all think is valuable and what we don't, and people are actually pretty good at it. But if you're not seeing results within three years, it's difficult to go raise the next VC fund, the next private equity fund, or just to raise more money for whatever your next fund vehicle is. And so what happens in practice is, people don't go spend their time investing in projects that are going to take a really, really long time and won't get marked. So what that means is, for an entrepreneur or for someone who's trying to get funding for something, getting that asset mark is unbelievably important because that's what lets the great investors go invest in you.   0:24:00 LC: So it's really important that for the VC company to get that Series B or the Series C or the Series D done. That single mark in time is hugely important because everyone can sort of concentrate on that, take it as a market price, even if it's not a perfect market price, and then write that in their books, measure it, sort of trust it to some degree, and everyone can sort of coordinate around that because you have a market clearing price there. And so if you think about it, just on the equity side of it, every founder's equity actually is a product in and of itself. I always find this interesting because I think most people don't think of it this way.   0:24:42 BR: I don't.   0:24:45 LC: But when you start a company, you're actually... You're selling two products. The first is sort of your individual product. This is the thing that you think you're starting. And the second is your company itself. And so your company can turn into a product where you sell your debt or you sell your equity or you sell some other sort of financing scheme, but that's a product, too. And the way that product is priced is, in the private markets, you have one-off auctions where you sort of game the options as much as you can to get the highest price. This is where everyone in their C, Series A, B, C... Well, not so much in C, but in A, B, C, you basically create auctions where you try to get all of the partner meetings on Monday morning to be talking about you, put all of your meetings into a week, and then you get everyone to bid all at the same time, and then you maybe don't go to the highest bidder, but you go with some mix of the highest bidder plus the people that you want to work with.   0:25:35 LC: Then the public markets are actually a totally different mechanism, it's a different distribution method where it's a continuous auction, where there's bids and asks continuous in time, at all times. And so you can't actually create these small little one-off auctions where you can rig the price up because the bids and asks, they just keep coming. But the benefit is, if you know how to... If you do well in that channel, you then have a lot of liquidity and you can usually get a higher price and arguably more capital. It's not actually even clear that you need to do that, but that's sort of the argument. And so I think if you start thinking about it that way, you can start to recognize, "Alright, that's why some projects are getting funded and some aren't." It's because the projects that are getting funded, they are products that work well in that market, and they are actually products, it's not just a throwaway phrase.   0:26:37 LC: I was chatting with someone about this earlier. I think it's probably good to take the emotion out of whatever project you're working on and think about this for unemotional things. So one of my friends is trying to get a research project funded, sort of like an arts VR research project funded. And we're talking about this and she's like, "Oh, now I get it. I should think about this like soap." So imagine you are a soap manufacturer, and you have made the best soap in the world. You think it's better than any other soap. You wouldn't expect to sell that just because you've created it. You'd think, "Okay, how am I going to get it out there? Am I going to get it on to Amazon? Am I going to start a store on Shopify? Am I going to go to the people at Costco or Walmart and cut a deal with them so it's distributed?” Because I might have the best soap in the world, but some mediocre soap that gets into the Costco channel and then works with those constraints, they are going sell more than I am. That product is going to do better. And if you care about people using your product and you're sort of not just cash flow-driven, but you actually care about the impact, you really, really need to think about that distribution channel and how you're going to get it out there.   0:27:50 LC: What you quickly find is that often the constraints that people place on their products, it's not that they don't realize they're making their products worse, it's that they want those products to get distributed and they think the tradeoffs are worth it. And so the really interesting new products, they recognize that, "Oh, there's some constraint or there's some tradeoff that a lot of other people made with their existing product lines, and I don't have to do that," because the way you distribute it has changed, or some assumption that they've made, they actually don't have to make that tradeoff. And I use something like soap because it's boring and unemotional to at least most of us, but it's almost definitely true with research funding. And so you and I talk about this a lot, but I mean, if I were trying to go raise money for research, it would depend what I was trying to do, but I think there are probably new distribution channels out there, so I mentioned with small scale... Sorry, you were saying?   0:28:50 BR: Oh, there's just three different directions that are really exciting to go with this.   0:28:56 LC: Oh, please.   0:29:00 BR: Yeah, so I think what I'm going do is I'm going to lay... Actually, I will lay out the places that I think are all tied into this that are all really interesting, and you let me know how you want to weave through them. So one is actually this... So both this point about a project as a product is a little bit mindblowing, and I think that it's tied to an earlier point that you made that I wanted to dig into about what it means to be legible to institutions. And if I am understanding correctly, the marking of valuations is one of the ways in which... At least, in the startup world, venture capitalists make themselves, their firm, as a product legible to other institutions. And so Shopify comes along, and you can now distribute your soap through an online store that you never could. What would be the project funding equivalence of that new distribution channel?   0:30:17 LC: So I absolutely don't think that this is that new, but it seems to have come somewhat in vogue, and I think it's just patronage. And so if I were trying to go do research where I was trying to make, say, call it $100,000 a year or something along those lines, basically enough that you could live a really good life, afford rent in any city and sort of have basically time to yourself, I think the obvious way to do it is to try to build an online following. And this is not a new idea. Kevin Kelly wrote that old essay, I think it was 1000 True Fans, where he said, “Look, at 1000 True Fans paying you $10 a month, that's enough.” I think a mutual friend, Andy Matuschak, who has Quantum Country has done a great job with his Patreon. I think it would be really, really, really difficult to do this. But I would think a lot about what really causes someone to say, "I'll pay $5 a month to go read this newsletter, or to go basically fund some research I find interesting." And this distribution mechanism didn't really exist before, and so I actually think in some ways, we're still pretty early on. And all I would do is think, "Alright, I need to get 2000 people to sign up all over the world." The Internet rewards niche behavior, and so how do I get into the community of these people find it just sort of interesting, and this is sort of entertaining to them, and I would think a lot about how I could create something around there.   0:32:01 LC: For the larger amounts, I would actually do the opposite. So for the larger amounts, I would go become friends with everyone in the funding world. So they have incentives too. And what you'd want to think through is normally... I guess I'll put it this way, and I was chatting with my friend about this. Normally, the way that the great researchers I know think, they're almost... They're quite dogmatic, to be honest. They say, "Okay, my project is the best project. This really will advance the field." But in practice, what might make it easier to sell the project is to understand what gets the person funding the project promoted? What makes the funder feel good?   0:32:40 LC: What will get to that next level of funding for the person above them too? And then if you're able to map that out, you can represent it in a way that basically works for everyone. And she was actually pushing back on me and saying, "Look, I don't want to lie. I don't want to represent my project that way. That seems sort of fake or it seems like a veneer." But the truth is, is that the project that she has in her head only exists in her head and doesn't exist in anyone else's head that way. And if she doesn't communicate it in a way that actually makes sense to them, then it's not going to get anywhere.   0:33:21 LC: So I think the really frustrating thing to come out of this is that basically everyone's in sales in some way, shape or form, and I think a lot of people don't want to be in sales or think that it is a sort of a difficult thing to go do. And so as a result, they just sort of shy away from it. And so this is, again, why I think the distribution analogies really, really can work well, because it sort of takes the emotional weight out of it. And then if you look at this and say, "Oh, this isn't the best grant maker in the world, this is just Costco, and I'm just trying to get into the new line," I think it can feel a lot less heavy. And you can maybe treat it, and maybe the field might open up to you a little bit more.   0:34:05 BR: Okay. I guess, the tension I see there is building up trust with the people who are the capital allocators, almost feels like the opposite of figuring out a different way of making yourself legible to an institution. Institutions are obviously made up of people, so these aren't two separate things. But I think that there's something to the fact that you need trust when you're doing something that is not institutionally legible. So it's like you don't actually have trust with a lot of the companies that are publicly traded that you invest in, but they are... They've packaged themselves in a way that is sort of institutionally legible if that's... And I think this might actually be a good point to really... What do you mean by something being institutionally legible? What does that mean?   0:35:20 LC: It's a vague handwavy way of saying you just need to be recognizable to the people who are buying your product, and you just have to understand, in practice, how those relationships work. And once you understand the practicalities of whatever market you're working in, then you'll be able to understand how to craft a product for the people who actually want it. And, again, I think the difficult thing here, this is not intellectually that challenging, it's much more of an ego thing where we have to put aside what we think are the best products that everyone should be buying or what everyone should be doing. So if you think about it, since we're talking very abstractly here, what capitalism really rewards is, and actually this is true of all non-violent selection, it rewards behavior change. And so what we're really saying is how do you get someone to sort of change that behavior. And when you think about it that way, what's legible in your head, if someone else hasn't learned all the same things you have, they're going to end up using some sort of abstraction, some sort of shortcut.   0:36:41 LC: And that's sort of what I mean by saying intellectually... Or sorry, institutionally legible, is you understand the abstractions they use, you understand basically the mental models they're using to try to understand what's going on, and then you are able to fit your product into that. So I can give you a couple of examples and findings that are...   0:37:02 BR: Yeah, please.   0:37:04 LC: So I don't know how deep into accounting you are, but there is a metric that's really commonly used called EBITDA. And effectively, it is a free cash flow proxy metric. And it was invented by some people in the cable industry who wanted to raise a lot of money to go roll out cable systems all across the US. And they wanted to be able to quickly raise debt to go buy these sort of small cable operators and then put them all together. And with this metric that they invented, all of these other investors suddenly had a Schelling point. Suddenly, all of these investors had a new unit of measurement to look at this type of business. And because they accepted it, they were willing to go fund those purchases. Suddenly, a whole wave of those purchases were done, and basically a whole wave of these projects were financed because someone figured out a way to make that institutionally legible.   0:38:11 LC: And a similar thing has happened in the last 10 or 15 years with what we call enterprise SaaS companies, where we now have a new set of metrics that weren't really in use 20 years ago. These are metrics, I'm not sure if you're familiar with them, these are metrics like...   0:38:26 BR: The CAC.   0:38:27 LC: Gross churn... CAC, gross churn, net dollar retention. And if I went to someone today and I said, "Oh, I'm investing in a business that has an average customer life time of six years, an LTV to CAC of 4:1, it has 98% gross retention and 127% net dollar retention, and I think those numbers are going to persist for the next four or five years, that is something that I almost wouldn't even have to explain what the product is. If something met those metrics and truly met those metrics, it's a company that would get a huge valuation in today's markets. And it's again, it's because it's now institutionally legible. Someone has basically convinced the world of that. So then the question should probably be, why do these things get institutionally legible? And what I find is that, we're actually re-using the same math over and over again and finding new situations where we didn't realize that math applied. And so usually what's happening is, we're finding relationships that are really durable, that are really, really, really resilient.   0:39:40 LC: So I have this little questions page on my website, and the first one is, "What is the next durable customer relationship that we haven't really seen yet?" So what happens is, once the market recognizes that there is a durable customer relationship, and you can build that into our models. These models actually should come from how we model these bonds that last 20 or 30 years. If you can fit the customer relationship into that model, suddenly, all of the bond investors and sort of the bond valuation metrics that we used as proxies, they drift into the financing world. And people say, "Oh, this is also a durable relationship, so we should go fund it." And coming back to your first question to say, how do some of these huge technology projects get off the ground, it's because someone has convinced a set of investors somewhere that there is this long durable, and that's important, resilient set of cash flow streams 20 or 25 years out, and then we discount that forward, so that's how that works.   0:40:45 BR: Oh, man. Okay, so to riff on that and to go back to your analogy to products and distribution channels, what basically... You could almost think of it as someone coming in being really good at sales and arguably like marketing, and basically changing taste and creating a new product category where people didn't know they wanted gluten-free things, and then they go and they create that new marketing category, and now customer tastes change, would that be...   0:41:29 LC: And it's funny you use the word taste, that is... It's both fundamental reality of, Oh, in a true Bayesian universal sense where we're updating our priors correctly, imagine we had all knowledge, that does matter. But then taste does matter too, that's exactly right. There's another book I'd recommend called "The Empire of Value" by a guy named Andre Orlean, who is this really interesting French economist. And so in this book, he makes this argument that prices are completely socially constructed, and it's like you're saying, it's taste. As a side note, it's totally unclear to me why all of the people who are coming up with the socially constructed value theories are all these French people. It makes one wonder what's in the water in Paris. But similar is to say, actually, I think, and everyone else thinks, and we're all sort of self-referentially thinking, therefore, the thing exists, the price exists, the value exists.   0:42:32 BR: Yeah, yeah, that makes sense.   0:42:33 LC: It exists as this organizing principle, which everyone else then cites as a real reference and then it takes on a momentum of its own.   0:42:44 BR: And what... And so, I guess, do institutional structures like C-corps and LLCs, do those relate to institutional legibility? In my head, they do, but I might be going a step too far.   0:43:04 LC: Yes, they do, but I want to backtrack in terms of what you're saying.   0:43:12 BR: Yeah, do it.   0:43:14 LC: So what they do is they basically... The legal structure sets the landscape for markets. I should completely confess my own bias here. I am massively, massively pro-markets. I think virtually, no other social mechanism that we know of has raised so many people out of poverty. But as much as I love markets, I recognize that it's not sort of this shallow teenager's love of markets where I overdosed on Ayn Rand. It's more of on the lines of...   0:43:45 BR: Be nice to the little libertarians.   0:43:49 LC: No, I was once one when I was 14 too, I get it. And I think the problem is, you have to understand markets are these amazing and emergent phenomena that pop up basically naturally everywhere, people trade with each other. But efficiently functioning markets are actually very, very expensive public goods to maintain. And that means that you're depending on the bias of all the regulators to try to make the best guess as they can to create and maintain these liquid markets to make sure that people are transacting fairly. To give you another book recommendation, there is an economist named Alvin Roth, who wrote a book called "Who Gets What and Why," and a lot of his students went on to go work at Uber and Airbnb to sort of create these marketplaces. And if you look at it, they're actually quite intentional about how they're sort of creating the markets. So now, let's take one step further back and say, “Alright, all of the countries are creating markets themselves, too, and they're creating the balance of these markets.”   0:44:54 LC: So as you know, I'm a lawyer and was a history major and sort of loved looking into this stuff. I would argue that one of the least appreciated social technologies of the last few centuries is the concept of limited liability. And so it used to be, before we had easy access to creating limited liability organizations, if you started a business and it went bankrupt, you personally went bankrupt. Maybe you were thrown in jail, maybe your family went bankrupt, and so you couldn't go that far out onto the risk curve. And so, socially, if you were thinking about this sort of like an agent-based modeling perspective, if you could basically increase the variance of what agents could do, if you could basically socialize some of the risk, then you let people take a little bit more risk. Maybe it doesn't work out as well for a few people, but socially, you get to that higher hill in the hill-climbing analogy. And so you're asking about how C corps work and LLCs work. Do you want me to just run through the history really quickly?   0:46:01 BR: Well, I guess more what I'm poking at is just talking about how, at the end of the day, these aren't laws of nature, the structure of organizations and...   0:46:14 LC: Not at all. So why do we have Delaware C corps? Coming back to limited liability, in the late 1800s, New Jersey created a charter that let anyone go get a corporation. And then after that, later in the 1800s, New Jersey passed a set of laws that are colloquially known as the “Seven Sisters,” And these were these terrible laws in the view of all the businesses who were registered there, so they were looking for other places to register. Delaware saw this as an opportunity, so around 1900, Delaware lowers their taxes, lowers their registration fees, and they bring a lot of corporate registrations in. And then they set up their court systems so that they specialized in registrations, at which point Delaware becomes the de facto place. You get a runaway phenomenon, then all of the good corporate lawyers want to go practice in Delaware or they want to be corporate judges in Delaware, and all of the interesting cases go to Delaware. And it's literally gotten to the point where everyone in the US references Delaware corporate law, and non-US companies will create charters saying they'll defer to Delaware corporate law, and countries who are still forming their legal systems will effectively copy and paste a lot of Delaware corporate law. And so coming back to your point, it's not a law of nature. These are people doing the best they can to optimize the landscape, and that's how it works.   0:47:47 BR: And so my thought would be that that does relate to institutional legibility, because if I went to someone and said, "I'm using a B corp structure," they'd be like, "What the heck is that? I'm not touching that with a 10-foot pole." But if I say that I am using a Delaware C corp, then that is a legible abstraction, so I guess that would be my argument for why institutional structures matter.   0:48:24 LC: They do, and I think what it comes down to is you have all these degrees of freedom when you're starting any organization or any project, and you just want to think about where you want to innovate and where you don't want to innovate. So you look at US business organizations, I should say this, since I'm a barred attorney, this is not legal advice. There are basically four options. You default into being a partnership where you actually have unlimited liability. You can be a limited liability company, which is done state-by state. You could be an S corp, which is a tax status of LLCs, or you can be a C corp, which is the one that you're talking about.   0:49:03 LC: And what you go see when you run through all of these things is, well, there might be a better way to do this, but for the company that I'm starting or the project I'm starting... So the fund that I run, we have a Delaware LLC. I could argue to you that there are things we could do that would actually be better for the investors and better for the whole strategy, but you then look at this and say, "Hmm, it's just not worth the marginal effort given the payoff of actually trying to overcome that sort of legibility hurdle." And so I think what ends up happening is you end up getting these innovations around the edges where someone says, "Okay, here is one use case that's a little bit better, and we'll keep everything else the same except for that," and then the new standard arises. I don't think it ends up being worth saying, "I want to create a new legal structure and a new product and do physics research all at the same time," just because there's not enough time in the day.   0:50:11 BR: Yeah, I guess it just... It makes me wonder, because it feels like these legal structures do impose certain constraints, it just makes me wonder out in the landscape on a completely different optimization mountain what other constraints could be imaginable.   0:50:40 LC: So probably the most difficult cost to measure out there is opportunity cost, because it's so difficult to say, what could things be if we organized everything differently? And one hopes that when you have 50 states, that's how federalism works in the US, one hopes you get people experimenting with regulation, and you can get maybe a new project started off the ground somewhere else, if not in the state that you live in, and then of course, with more countries, you can maybe go overseas and do it too. And it's interesting, you brought up Spotify a little bit earlier, it's unclear to me that Spotify could have gotten started in the United States, given the state of music laws at the time. But then what happens is all of these European customers start using the product, and that has an institutional legibility of itself, and people say, "Oh, okay, I can see it's working in that country, it will probably work here," and I wasn't involved in the record label negotiations, but I assume that's basically what they were looking at. And then you look and say, "Oh, okay, then the laws can change."   0:51:52 LC: The other thing that I just want to point out is that when a law is set, that's a much more fluid thing than I think most people realize who haven't spent a lot of time looking at this. So in practice, a lot of times, there are sort of these gray areas of the law, and I'm not saying people should go break the law. But there's a gray area of the law where the products that you're working with don't really fit into the regulation, or customer demand is just so massive that the regulators will actually change their mind once they see that demand. Now how far you want to push that boundary is really up to you. There are arguments that Uber or Airbnb were illegal when they were first started. There are arguments that they're illegal right now. I don't think so, and I think they did the right thing, and I think the world's a better place for giving everyone the options. But it’s also really, really important to realize is there are these constraints, but the constraints, when you read a law, it's not a law of physics. And the other thing that you have to understand is laws are executed by regulators, so understanding why they are enforced or what they actually want to enforce is also really, really important.   0:53:09 BR: Yeah, and do you think there's... So to your point about there being different regulations in different places, do you think that it's then problematic that you see so much copying of Delaware law and sort of copy-pasting that around the world? 'Cause wouldn't that then sort of make everything... Wouldn't that be a very strong attractor?   0:53:37 LC: I think what ends up happening is it's a good enough baseline. So I can't remember what the book is called right now, but there is another famous economist named Hernando de Soto who wrote about just the importance of property rights and how if you are able to sort of import the property rights regimes from the US into a lot of different countries that don't have them right now, it would be a huge driver [0:54:00] ____.   0:54:00 BR: It wouldn't necessarily work.   0:54:01 LC: And so I don't think we live in a world where we figured everything out so perfectly that all we need to do are these sort of minor experiments. I think we live in a pretty uneven world where if we just had relatively good legal functioning across the world, not just in terms of the laws that are written down, but sort of culturally how they're practiced, we could make life a lot better off for a lot of people. So it does make a lot of sense to me that if you and I were trying to start up a corporate law and corporate practice in some small country somewhere that was just starting to figure it out, or just decided they really wanted to change their system, I think we would go look at best practices. I think that's normal. It's unclear to me though that we are actually doing enough experimentation on the regulatory side, it's just really, really hard to say how much because it's just sort of this abstract opportunity cost question.   0:55:03 BR: Yeah, it's... And I guess these are sort of the same thing where I think of it as it's very hard to talk about counterfactuals, and actually, to riff off of the point about opportunity costs, my impression about... Of one of the reasons that large long-term projects don't get funded is because the opportunity cost is so high in that if I see that the stock market is increasing at a... It's like the number in my head is 5% of... I think of stock market 5%, I'm not... Is that roughly...   0:55:47 LC: I think nominally, the numbers, depending on the timeframe you look at, are along the range of 8%-10%.   0:55:56 BR: Oh, wow, okay.   0:55:57 LC: But there are actually a lot of people who right now think that 5% is what you're going to see for the next 10 years.   0:56:03 BR: Okay, well, let's...   0:56:05 LC: Anyway, doesn't really matter. Let's say 5%.   0:56:06 BR: Yeah, exactly. So in order to make the argument for something like the opportunity cost of investing in an illiquid thing is the compounding returns that you would get from 5% growth in the stock market, plus the amount that... Like the liquidity that you're giving up, which is, as you pointed out, a really big deal. And so it's... And then put uncertainty on top of that, so it's not even a guaranteed in the future compounding... Like you need to be... So it just... It seems like it's a fairly straightforward... It's actually a very, very large opportunity cost to propose an alternative investment to just the stock market.   0:57:07 LC: So I think it is and it isn't. First of all, I think you framed all of that correctly, that everything is subject to an opportunity cost. And so, of course, when I'm looking at whatever investments I'm making, and you are too, or deciding where to spend your time, you're going to look at your other alternatives and then choose. I don't think that necessarily should mean that it's impossible to go find a project worth working on. I think what it means is you just need to really, really understand what you're building. So that you understand why it's really valuable, and you have to go after sort of basically big projects or you have to have really, really fast experimentation, so you can just try out a lot of things and say, "Okay, maybe the opportunity cost is high over five to seven or eight years or 10 years, but I am going to try 2,000 different types of Shopify stores programmatically, I'm going to figure out which ones work, and then I'll have the revenue stream that I want once I've tested out and pulled out to the best 25, and then go on from there."   0:58:16 LC: So I do think that that it's definitely doable, you just have to recognize the opportunity cost. But you're right, there is an opportunity cost. I just think you shouldn't sell yourself short. I think implicit in what you're saying is that the world is relatively efficient, and because the world's relatively efficient, how on earth could I earn more than 5%? But I have to say, I look around everywhere and see a lot of products that, they were built on the constraints of the past distribution channels, they were built on the constraints of the past production approaches, or they were built on social relationships that have broken for whatever reason.   0:59:04 LC: So you look at this and say, huh, I think there's probably a better way to do it. And if I'm right, and if I really, really focus on figuring out what's wrong and how we can do this better, you're going to find that the returns you earn are massively more than the stock market. I just think you have to be really focused and intentional in how you're doing it, and I think you have to spend a lot of time understanding the people behind the process. If you ever... I'm trying to think. Have you ever read that essay "I, Pencil" by Leonard Read? There's this idea that if you look at any sort of product in front of you, so you look at a pencil, an uncountable number of relationships went into building that product. So for the pencil, someone had to chop the wood, someone had to mine the metal, someone had to refine it, someone had to put it all together, someone had to paint it, someone had to build the eraser, and someone had to invent all of that and patent all of that, and start all of those companies and then figure out how to market it, and then figure out how that distribution channel worked, and then figure out how consumer tastes were changing, and just look through all of that.   1:00:11 LC: There are so many relationships there, and if you think about it, there's just... There's no... It's extremely unlikely that we've reached the global maximum for almost any product, because you only need one of those relationships not to have been done perfectly, not to have been optimized, to have an opportunity to do things better. And then you look at the constraints that they used to have 80 years ago versus what we have now... Software has changed so much in the last 15 or 20 years, the Internet has changed the world a lot in the last 20 to 30 years. You look at this and say, there probably are better ways to organize these things or to sort of optimize things. And I think that's true... I'm looking around my apartment now, when you look at, I don't know, a glass, or you look at a countertop, or you look at any art or any hardware, I actually think this is true for almost the most mundane object in your life.  And actually I find... Once you start getting into the details of all of these mundane objects, it's not mundane, it's totally...   1:01:19 BR: My concern is actually the opposite, where I think that there are tons and tons of dollar bills on the ground, but the payoff you need to convince someone of becomes inordinately larger, the better the stock market is doing, it feels like, because of the opportunity cost.   1:01:45 LC: So yes and no. If you look, for reasons that are separate from this conversation, at demographics and the way that capital is structured, interest rates are low and look like they're going to stay low for a while, which means the required return for a project is going to keep falling. So yes, when the stock market is doing really well... Imagine the stock market were returning 40% a year, it would be harder and harder to get new projects funded because people would just put their money in the stock market. But as those returns fall from 8% to 5%, or you used to be able to get 6% or 8% over a 10-year period in a 10-year bond and now you're getting 2, 2 1/2% a year, you actually are more and more willing to go out onto that risk curve and sort of fund something new. So I actually don't think the problem is as much opportunity cost, especially today. Socially venture capital is so popular that I don't think the problem is opportunity cost. I actually think the problem is alpha. And so if you think about what alpha is in the finance world, it's basically, you're looking for an information advantage, and it's going back to cash flows and capital flows.   1:03:07 LC: You're looking for an information advantage on what's going on with those cash flows, with the product, the customer sort of thing, or what's going on with capital flows. So your alpha could be, you understand there's going to be a forced seller here or a forced buyer there, and then you bridge the liquidity into that market. And to throw one more book out there, the best book I know of to think about information sourcing is a book by a Nobel Prize-winning physicist named Robert Laughlin, it's a book called "The Crime of Reason." Have I ever mentioned this one to you?   1:03:39 BR: No.   1:03:39 LC: So it's really interesting. Frankly, it's a shocking book when you really process it. He basically argues that all economically valuable information is kept secret. And so you think that you really understand a lot about the world, but you actually understand, say, 98% about some topic, but that last 2% that really matters to get the project off the ground, to get the product built, to actually get funded, that's really kept secret. So the reason I think this is interesting is we've turned an opportunity cost problem of, "Well, there's really nothing I can do about it, I hope I come up with a good idea," to an information sourcing problem. So the way I think about this is I say, okay, there are really two places that you find information in the world. It can either be recorded or it can be in someone's head. So recorded could be like written and natural language, or in numbers in a database. And I often find, unless we're talking about you going and coming up with some new fundamental algorithm, all you really need to be doing is collecting all of that data and joining tables. It's not actually that complicated from an intellectual perspective, but it's really about finding those tables and joining them. And then on the side of, oh, it's in someone's head, it just ends up being about building relationships with people.   1:05:01 LC: And to your point about there being lots of dollar bills on the sidewalk, there are, but it's almost like they're invisible, so you need to go find the information to really understand, oh, that's a real one, that's a fake one. And it just ends up being a shoe leather exercise where you say, "Okay, I'm just going to go reach out to a lot of people, become friends with a lot of people, talk to them about their work, really try to understand what they're going through, and then I'll recognize what they want and what they don't want, and then I'll find effectively that alpha." And I think that's probably a more useful way to think about it than opportunity cost, because it's more empowering once you think about it that way.   1:05:38 BR: I like that. To change tracks a little bit drastically, but to just get to a point that I think it is really important to talk about... So you invest primarily in public equities, right?   1:05:52 LC: Mostly public, but public and private companies, yeah.   1:05:55 BR: Yeah, and so there is a argument that... I'm on like... There's basically an argument that short-term is like short-term thinking on the part of public investors has sort of pushed public companies to slash R&D costs and basically caused the fabled death of corporate labs. I think it is pretty clear that corporate labs don't sort of have the sort of world-changing output that they used to. However, I'm agnostic about the cause and still trying to figure that out, so... What do you think about that argument?   1:06:42 LC: So, I think it's complicated. I also think I'm not sure, but I can think it through with you.   1:06:50 BR: Yeah, let's think through it.   1:06:51 LC: Sure, so if you look at the valuations in the public markets today, they are very high by any historical measure. And so high valuations do not imply short-termism. They imply that the market is placing very, very high prices on companies. Now, it just turns out that a lot of that has to do with the way capital flows work today, not just with cash flows. And so what's going on effectively is we changed the retirement system in 2005. So we default decided to put a lot of people's money into index funds. Index funds just blindly buy a set of equity as a set of stocks, just as capital flows into them, and so we've had more and more retirement flows, so you see all of these stocks get bid up. That has been a huge reason for valuations going up. But anyway, you look at this and say, alright, so just on a project basis, companies are actually getting huge valuations. Now quarter to quarter, companies face unbelievable pressure to sort of make a mark that Wall Street thinks is good or bad. And what ends up happening is people are definitely optimizing over the quarters, because the research analysts, it's so difficult to see inside the companies that these are the metrics that they use to measure what's going on.   1:08:13 LC: So it's sort of a mixed bag. We are getting really high valuations, but there is still a lot of quarter to quarter pressure, but at the same time, I mean I look at this and say... I think it's actually closer to the journalism and editorial arguments, where it used to be that these newspapers were monopolies and then separately or sort of for social reasons, they were also safeguarding these unbelievable journalists, and it was this huge benefit to society. The reason it worked was the newspapers were monopolies, so they really didn't face competition, and then culturally it became normal for them to sort of support journalists. And then it was like a social competition, like "Who is going to win the Pulitzer price this year? Who's sort of funding the best journalists?" If you go look at the big corporate R&D labs, you find that it was a set of funders that were basically semi-government entities. They were such great monopolies, and culturally, the people who are running those companies also wanted the R&D labs, maybe out of the sense of patriotism, maybe out of some other sense, but I think that's sort of how they came to be. And when those monopolies were broken up, they basically weren't able to keep funding the R&D labs.   1:09:40 LC: I do think that some of today's monopolies and oligopolies, these are the Facebooks and the Googles and the Microsofts of the world, they are able to fund big R&D labs, and we could argue about whether it's the same as Bell labs or PARC... But they're definitely trying, they have been inspired by those old examples, and my friends who work there, I do think are quite brilliant. So I do think that the ones that you're talking about and that I've read that you've written about, I think that it was basically this really nice side effect of monopolies that also doing it. But at the same time, not every monopoly... And in fact, almost every monopoly isn't going to have that cultural imperative. And then on the flipside, let's look at the ones that aren't monopolies, and this is again, partially a narrative problem and partially a reality problem. People haven't come up with a good metric for outsiders to know that research projects are going to do well long-term, so the outsiders feel comfortable funding them.   1:10:48 LC: So an example is that over the last 15 years, you can go look at pharmaceutical companies, and you'll see that their R&D budgets are getting cut. And what happened was a lot of investors were looking at the returns on that R&D over a three-year basis and a five-year basis, and they were saying, "Look, we're not seeing any returns here, it really doesn't make sense for you to be spending money." And of course, people trot out the worst examples when they're making arguments, but there was a set of pharmaceutical companies that maybe was abusing the R&D line. Maybe they were basically not really doing great research, and they were paying themselves a lot of money to not do great research. And some hard-charging Wall Street hedge funds came in and really, really pressured those companies to stop spending on R&D. Now, you'd say socially, that's terrible outcome. We could say, "Look, maybe the R&D is a public good, not a private good, so we need some way to incentivize that and we can have that conversation." I think it's possibly solvable if we come up with a new set of metrics that everyone actually believes.   1:12:07 BR: Yeah, so this goes back to the legibility point.   1:12:09 LC: It does. So you and I have spoken about this one privately before, but there's a professor at MIT named Andrew Lo who proposed that you bundle cancer research projects together or any pharmaceutical projects together. And say you take 100 of these projects or 200 of these projects, you bundle them together, you give each of them, say, a couple million dollars, and then you bundle all the payoffs together. And so the idea is that, hopefully, that's institutionally legible enough that someone would be willing to fund it because they think, "Okay, there's actually a good chance that of these 200 projects, one or two of them will hit, and then you'll have this unbelievably valuable drug that will really be good for the world, and maybe that's a good way to push us out on the risk curve." I haven't seen this type of thinking really take hold because we're still very much in that project-based milestone-based financing approach where it's like, okay, you have the metrics that makes sense for your series A, for your series B and C and D.   1:13:18 LC: And there's also an argument that maybe the smartest biotech investors and pharma investors are already cherry-picking the best companies, the best projects. So maybe you'll sort of have this adverse selection where maybe of the top 200 projects, this would have worked, but if the best five are just going to go off on their own, you're just not going to get the good ones. And this is again, sort of that information s

Is trend following too big? Can managed futures do it without the bond tailwind they've had for 30 years? Does globalization take away diversification? We're debunking these trend following myths/truths (?) in today's podcast – and today's guest is ideal to take us through the ins-and-outs of these trend following theories. We're joined by Kathryn Kaminski, PHD, CAIA, and Chief Research Strategist & Portfolio Manager at AlphaSimplex who has written the literal book and research papers on these theories and more. We're also talking with Kathryn about AlphaSimplex, COVID puppies, the Nashville predators, diversifying across trends, Dr. Andrew Lo, research papers & books, pure risk premium, crisis alpha, trend following “doesn't work” theories, alternative data, being an MIT professor, homemade Swedish meatballs, risk/volatility targeting, and being an alternatives person in a stock town. 00:00-01:43 – Intro 01:44-13:18 = An Impressive Background with a touch of Sweden 13:19-35:03 = Alpha Simplex, Trend Models, & Crisis Alpha 35:04-53:19 = Debunking Trend Following's Dead & Inflation Environments 53:20- 01:09:51 = Risk Targeting & 2020: A year in review 01:09:52-01:13:29 = Favorites Follow along with Katy on LinkedInand check out the AlphaSimplex website. And last but not least, don't forget to subscribe to The Derivative, and follow us on Twitter, or LinkedIn, and Facebook, and sign-up for our blog digest. Disclaimer: This podcast is provided for informational purposes only and should not be relied upon as legal, business, or tax advice. All opinions expressed by podcast participants are solely their own opinions and do not necessarily reflect the opinions of RCM Alternatives, their affiliates, or companies featured. Due to industry regulations, participants on this podcast are instructed not to make specific trade recommendations, nor reference past or potential profits. And listeners are reminded that managed futures, commodity trading, and other alternative investments are complex and carry a risk of substantial losses. As such, they are not suitable for all investors. For more information, visit www.rcmalternatives.com/disclaimer

My guests today are Andrew Lo and Jon Gordon. Andrew is a Charles E. and Susan T. Harris Professor at the MIT Sloan School of Management and director of the MIT Laboratory for Financial Engineering. He is the author of Hedge Funds and the coauthor of A Non-Random Walk Down Wall Street and The Econometrics of Financial Markets (all Princeton). Jon is an American author and speaker on the topics of leadership, culture, sales, and teamwork. He has worked with numerous athletic organizations, academic institutions, and corporations. He holds a Bachelor of Science in human ecology from Cornell University and a Master of Arts in teaching from Emory University. The topic is Jon's book The Energy Bus: 10 Rules to Fuel Your Life, Work, and Team with Positive Energy. In this episode of Trend Following Radio we discuss: Market Data Human Behaviors Financial Market Market Crash Adaptive Market Hedge Funds Strategies Market Dynamics Jump in! --- I'm MICHAEL COVEL, the host of TREND FOLLOWING RADIO, and I'm proud to have delivered 10+ million podcast listens since 2012. Investments, economics, psychology, politics, decision-making, human behavior, entrepreneurship and trend following are all passionately explored and debated on my show. To start? I'd like to give you a great piece of advice you can use in your life and trading journey… cut your losses! You will find much more about that philosophy here: https://www.trendfollowing.com/trend/ You can watch a free video here: https://www.trendfollowing.com/video/ Can't get enough of this episode? You can choose from my thousand plus episodes here: https://www.trendfollowing.com/podcast My social media platforms: Twitter: @covel Facebook: @trendfollowing LinkedIn: @covel Instagram: @mikecovel Hope you enjoy my never-ending podcast conversation!

Andrew Lo and Jon Gordon Interviews with Michael Covel on Trend Following Radio.

On this week's episode of The Freenoter, we look at TED/TEDx talks - are they worth it for freenoters? What makes a great TED-style talk, and why they may or may not be right for you. We also look at one of the TEDxCambridge speakers Tamsen worked with, Andrew Lo, and what goes on behind the scenes during a TEDx event. And we debut an original-ish cocktail, the tequila-based Spectre. Show Notes: Tamsen's TEDx Talk Flowchart ("Should I do a TEDx talk?"): https://tamsenwebster.com/tedx/ Andrew Lo's TEDxCambridge talk: https://www.youtube.com/watch?v=xu86bYKVmRE Great continuity errors from Hollywood: https://screenrant.com/worst-continuity-mistakes-movie-history/ Toilet Injuries: https://www.answers.com/Q/How_many_toilet_injuries_in_US Drew Tarvin: https://drewtarvin.com/ D'Angelo, "Untitled": https://www.youtube.com/watch?v=SxVNOnPyvIU Jill Bolte Taylor: https://www.ted.com/talks/jill_bolte_taylor_my_stroke_of_insight And The Spectre: 2 1/4 oz. good tequila (we used Don Julio Blanco) 3/4 oz. Cointreau pinch of table salt Stir over ice in a cocktail mixer and strain into a coupe. Slice off a thick, coin-sized hunk of lime rind (catch a little lime flesh with it) and squeeze it over the drink. Remember not to be fooled by Le Chiffre's twitching eye.See omnystudio.com/listener for privacy information.See omnystudio.com/listener for privacy information.

This episode features Prof. Andrew Lo, the author of a paper that we discussed recently on Linear Digressions, in which Prof. Lo uses data to predict whether a medicine in the development pipeline will eventually go on to win FDA approval. This episode gets into the story behind that paper: how the approval prospects of different drugs inform the investment decisions of pharma companies, how to stitch together siloed and incomplete datasts to form a coherent picture, and how the academics building some of these models think about when and how their work can make it out of academia and into industry. Professor Lo is an expert in business (he teaches at the MIT Sloan School of Management) and work like his shows how data science can open up new ways of doing business. Relevant links: https://hdsr.mitpress.mit.edu/pub/ct67j043

When it comes to Behavioural Finance, a few people stand out in terms of their contribution to helping us all understand why and how it works. The intersection between Human Behaviour and Quantitative Investing can be difficult to understand for even the most sophisticated investors.  Today, I want to share some really important insights from one of my favorite professors, who is also a practitioner of this discipline, namely Andrew Lo of MIT Sloan School of Management and Director of MITs laboratory of Financial Engineering.  Many people know Andrew as the father of the Adaptive Market Hypothesis, and our conversation was wide ranging, entertaining, and deeply insightful.  So enjoy these truly unique take aways from Professor Andrew Lo, and if you would like to listen to the conversation in full, just go to Top Traders Round Table Episode 18 and Episode 19. The Ecosystem of the Financial Markets Niels: Now, our conversation today will focus on a number of different topics within the managed futures industry, and, perhaps, a few that will fall a little bit outside of this. So, to kick things off in a slightly different way, I want to come to you, Andrew, first, and ask what you think of when I say Rabbi Mahony, Rabbi Mahony, Rabbi Mahony, and I hope you know what I’m referring to so that our listeners don’t think that I’ve completely lost it at this stage. Andrew:  Well, thank you for bringing that up. That comes from one of my stories that I’ve written about in my book, Adaptive Markets. It’s an idea about thinking about financial markets more like a biological ecosystem rather than a physical system.  As you may know, most economists suffer from this disease that I call physics envy. We wish that we had three laws that explained ninety-nine percent of all behavior, the way the physicists do. In fact, we have ninety-nine laws that explain only three percent. So, the idea behind adaptive markets is that we really have to think about these financial market dynamics as coming from human interactions, and trying to model those human interactions is really critical. So, the Rabbi Mahony story really has to do with the fact that I heard many years ago about a technique for getting parking in Harvard Square.  It’s a terrible, terrible challenge to drive a car into Harvard Square because there is never any parking. So, for years I just decided not to do it. But, a friend of mine said that, if you used this following algorithm: before you go to Harvard Square you utter the incantation, Rabbi Mahony, Rabbi Mahony, Rabbi Mahony, at that point you should be able to go to Harvard Square and get parking. The amazing thing is this algorithm actually works. But, the more interesting reason is why it works. It works because it changes the way we behave. It changes our expectation for getting a space. Because now, once you utter the incantation, you must, somehow, in a part of your brain, believe that you might be able to get a space and that changes the way you drive. It changes how you look for parking, and, magically you actually increase the chances of getting a space. So, it really says that human behavior can actually change our reality.  Sometimes things need to be believed in to be seen. Niels:  Yeah, absolutely. Just out of curiosity, do you think that belief always precedes action and plausibility? Andrew:  I think it is something that happens simultaneously, in many cases. Our beliefs have an impact on our behavior, but our behavior has an impact on reality, and that reality shapes our beliefs. So, it’s kind of a feedback loop that is happening and updating all the time. Unless we’re aware of that, it’s very easy for us to get mislead by various kinds of market events and ultimately end up down a rabbit hole of behavioral biases that ultimately end up hurting us in our investment strategies. Niels:  Yeah, well I look forward to finding out whether this little chant also works finding a parking place here in Switz...

with Andrew Lo (@AndrewWLo) and Jorge Conde (@JorgeCondeBio) The advent of new gene and cell therapies are beginning to approach that holy grail of medicine—that of a possible cure. But they are also more expensive than any medicines ever sold before. In this episode, MIT economist Andrew Lo and a16z General Partner on the Bio Fund Jorge Conde discuss how exactly we place an economic value on a cure; the questions we still need to figure out, like who should pay for what and how; and how we need to start thinking about handling the coming influx of highly priced medicines like these into our healthcare system. If we think about these payments as a kind of 'mortgage for a cure,' what happens when your gene therapy mortgage defaults? How would payment plans like these move between insurance plans? Lo and Conde also discuss the broader context in our healthcare system, the economics and risk of drug discovery and development overall – and finally, how our markets might just function more like biological systems than anything else.

"If someone wants a better return, they have no choice but to take additional risk, no matter how they feel about risk or how risk averse they may be." - Sol Waksman (Tweet) Welcome to Top Traders Round Table, a podcast series on managed futures brought to you by CME Group, where host Niels Kaastrup-Larsen continues his conversation with Andrew Lo, the Charles E. and Susan T. Harris Professor at the MIT Sloan School of Management, and Sol Waksman, the Founder and President of BarclayHedge, Ltd. Listen in to learn the effects of politics in the financial markets and how trend following fits in, the state of cryptocurrency in the current market, and the advice our guests have for investors and their future. Subscribe on: In This Episode, You'll Learn: Why relatively so few investors have added managed futures and trend following to their portfolio, despite all the evidence The lessons different groups of people can still learn from the economic crisis of 2008 How political uncertainty affects where investors put their money "[Many of these finance] technologies are now creating new sub-industries. Who would have thought that cryptocurrencies would be a separate asset class, but it seems like it's emerging as such." - Andrew Lo (Tweet) What Sol sees as the constant lesson we should be learning from these periodic economic events Why investors have a hard time grasping the advantages of the liquidity that trend following brings to investments The present and future impact of artificial intelligence on the finance industry "What have we learned from this last stock market crash? I think we keep learning the same lesson, and that lesson is that when liquidity dries up, all correlations go to 1." - Sol Waksman (Tweet) Why Andrew believes the centralization to one cryptocurrency is inevitable The advice Andrew and Sol have for investors to prepare for the future This episode was sponsored by CME Group: Connect with our guests: Learn more about Andrew Lo and MIT Sloan School of Management Learn more about Sol Waksman and BarclayHedge, Ltd. "I think there were a lot of lessons that were offered by the financial crisis, but the real question is who actually took those lessons to heart." - Andrew Lo (Tweet)

"I was going through one of those inevitable losing streaks [early in my career] and I had a fundamental question. Does anyone make money trading the futures markets, or is it all a casino set up for the benefit of the exchanges?" - Sol Waksman (Tweet) Welcome to Top Traders Round Table, a podcast series on managed futures brought to you by CME Group. On today's episode, host Niels Kaastrup-Larsen speaks with Andrew Lo, the Charles E. and Susan T. Harris Professor at the MIT Sloan School of Management, and Sol Waksman, the Founder and President of BarclayHedge, Ltd. Our guests today have many years of experience in the world of  investing and have watched the finance industry as it has gone through it's many changes. Listen in to learn the historical importance of separating the alpha from the beta, the increase of the volatility of volatility since 2007, and common misconceptions about diversification in a portfolio. Subscribe on: In This Episode, You'll Learn: How our guests got their starts in finance Why Andrew thinks that sometimes things need to be believed to be seen Why separating the alpha from the beta has been revolutionary for the entire financial industry "Markets are efficient most of the time. Every once in a while, human behavior ultimately overwhelms the kind of rational deliberation that efficient markets are based on, and we do see periods of fear and greed that ultimately take over. But it's not one or the other, it's both." - Andrew Lo (Tweet) How the narrative of finance and investing has changed over time in the futures industry The increase in the volatility of volatility over the last decade and what that means for managed futures "I think that for the most part, investors misrepresent what they want." - Sol Waksman (Tweet) How the abundance of passive investing tools has changed the investor's experience during market changes What investors are looking for in their tools and strategies What alternative strategies are investors looking at now "The one lesson we learn from academics throughout the whole process of passive investing is that diversification is really key." - Andrew Lo (Tweet) The experiment Andrew does with his audiences he speaks to on their investment preferences What investors need to understand about diversification and risk This episode was sponsored by CME Group: Connect with our guests: Learn more about Andrew Lo and MIT Sloan School of Management Learn more about Sol Waksman and BarclayHedge, Ltd. "A stock index has no relevance to the performance of a CTA, but if you calculate alpha by regressing against the stock index, you have a number that is absolutely meaningless." - Sol Waksman (Tweet)

Customer and employee safety, risk assessments and policy recommendations - machine learning is fast becoming a fundamental part of the financial services ecosystem. Hosts Jodie Wallis and Amber Mac examine AI's evolving role in financial services, particularly within insurance - featuring interviews with John Heveran of Liberty Mutual, Tomi Poutanen from TD's Layer 6 AI, and former CEO of Kanetix Ltd. Andrew Lo. If you enjoyed this episode, please rate and subscribe to The AI Effect on your preferred podcast app. To learn more about the topics covered in this episode, go to our website, theeffect.ai or accenture.com.  Follow us on Twitter @AIEffect.

MIT quant says next project will be to combine behavioural science with tech such as machine learning

Economist Andrew Lo talks to the FT's John Authers about his adaptive markets hypothesis, the idea that markets develop and adapt over time and should be modelled using concepts from biology instead of physics. It's the subject of his recent book, Adaptive Markets: Financial Evolution at the Speed of Thought. This interview was originally published on March 24, 2017. Music by Podington Bear. See acast.com/privacy for privacy and opt-out information.

There are two popular schools of thought with regards to how markets work. There's the efficient markets hypothesis (EMH) which says that it's basically impossible to beat the market, because all information is completely priced in at all times (more or less). On the other side is an increasingly popular behavioral view which argues that various human emotions and biases are always creating situations that aren't justified by the data. On this week's episode of the Odd Lots podcast, we speak to Andrew Lo, a professor of finance at the MIT Sloan School of Management about his own theory, which he calls Adaptive Markets. The theory attempts to bridge the behavioral approach with the efficient markets view. He argues that the proper way to view the market is through an ecological lens, examining the players as flora and fauna of a complicated system, to help determine who's thriving, who's dying, and where asset prices will go.

Andrew Lo, Professor at the MIT Sloan School of Management and Director of the MIT Laboratory for Financial Engineering, joins hosts Jeremy Schwartz and Jeremy Siegel to discuss his new book, "Adaptive Markets: Financial Evolution at the Speed of Thought" with a recent market update on Behind the Markets. See acast.com/privacy for privacy and opt-out information.

Guest: Andrew Lo - the Charles E. and Susan T. Harris Professor at the MIT Sloan School of Management and director of the MIT Laboratory for Financial Engineering. He’s going to talk to us about his book Adaptive Markets: Financial Evolution at the Speed of Thought, which has been shortlisted for the 2017 Financial Times and McKinsey Business Book of the Year Award See acast.com/privacy for privacy and opt-out information.

How financial markets work has long been the subject of debate between two academic camps: the efficient markets theorists and the behavioral economists. Into the breach stepped finance professor and author Andrew Lo, who joins Matt to discuss his efforts to bridge this ideological divide. Instead of viewing financial markets as “a physical system like a mechanical clock,” says Lo, “we really need to look at it as an ecosystem with particular organic agents that are acting with each other.” The implications of Lo’s work extend far beyond the ivory tower, to regular investors and society at large. Andrew W. Lo is the Charles E. and Susan T. Harris Professor at the MIT Sloan School of Management and director of the MIT Laboratory for Financial Engineering. Your host, Matt Miller, is the policy and communications advisor for Capital Group. An author and former Washington Post columnist, Matt was co-host of the public radio program Left, Right & Center. Do you have any topics for Capital Ideas? Please contact our editorial team at CapitalIdeas@capgroup.com. Related: (In U.S.) 4 Behavioral Tips to Help Investors Avoid Emotional Self-Sabotage In Europe: https://www.capitalgroup.com/europe/capitalideas/article/behavioral-economics-tips.html In Asia and Australia: https://www.capitalgroup.com/apac/capitalideas/article/behavioral-economics-tips.html The Capital Ideas websites are not intended for use by Canadian audiences. In Canada, please visit thecapitalgroup.com/ca for Capital Group insights.

To celebrate the FT’s Business Book of the Year Award, the team talk to the six shortlisted writers. In this fifth episode, Isabel Berwick, assistant features editor, and Andrew Hill, management editor, hear from Andrew Lo, author of Adaptive Markets, on his hypothesis that the theory of market efficiency is not wrong but is incomplete. See acast.com/privacy for privacy and opt-out information.

Review of Andrew Lo's "Adaptive Markets" (May 2017). The book expands the Efficient Market Hypothesis to account for times of irrationality.

The family of Wan Li Zhu did not see a future in China. His parents, persecuted by the one-party state, came to America when Wan Li was ten years old. China’s great loss became America’s brilliant gain. Wan Li benefited from high-quality public education at Bronx High School of Science and went on to a perfect grade-point average at MIT. He studied under renowned quant wiz Andrew Lo and was poised for a career on Wall Street but was lured away by the prospect of hands-on responsibility for product features at Microsoft. After a prodigiously successful stint, during which he was involved in building and marketing Dynamics CRM, MS’ fastest-growing product, he went to Harvard Business School. From HBS he was recruited by early-stage VC firm Fairhaven Capital. The firm, known for its expertise in web security and digital advertising, now sees promise in various applications of artificial intelligence starting with self-driving technology. Wan Li is deeply engaged in bringing on the next generation of winning investments at Fairhaven Capital. Despite a busy professional life, Wan Li Zhu has found time to advise startups and to co-found MIT Angels in Boston. I learned a ton from my conversation with this wise, yet unassuming VC. Here is a list of some of the topics broached: Wan Li Zhu Bio From Persecution in Communist China to Bronx High School of Science Studied with MIT Professor Andrew Lo – Used Natural Language Processing to Assess Market Sentiment Why Wan Li Zhu Went to Microsoft – Three Years at MS – Shipped Three Versions of the Product Wan Li Zhu Connects with Fairhaven Capital through HBS Resume Book Fairhaven Capital Is Thesis-driven – Attentive to Market Trends that Could Create Large Opportunities How the Fairhaven Capital Portfolio Is Doing What Wan Li Zhu Looks for in a Startup Investment Experienced Founders Can Actually Time Markets TVision Came Via MIT Angels – Measuring Engagement of TV Viewers AirFox – Enabling Wireless Carriers to Offer More Affordable Data Plans MIT Angels Company PathAI’s Deep Learning System Is Better at Detecting Tumor Cells than Human Pathologists Latch – Enterprise-grade Keyless Access System for Apartment Buildings The Investment Wan Li Zhu Regrets Not Making Wise VC Wan Li Zhu Continues to Be Very Bullish on AI

My guest today is Andrew Lo, the author of “Adaptive Markets: Financial Evolution at the Speed of Thought.” He is also the Charles E. and Susan T. Harris Professor of Finance at MIT and the chairman and chief investment strategist of the AlphaSimplex Group. The topic is his book Adaptive Markets: Financial Evolution at the Speed of Thought. In this episode of Trend Following Radio we discuss: Efficient market hypothesis Adaptive markets hypothesis The random walk hypothesis Crowded trade phenomenon 2008 meltdown Paul Samuelson Commodities Corporation Jump in! --- I'm MICHAEL COVEL, the host of TREND FOLLOWING RADIO, and I'm proud to have delivered 10+ million podcast listens since 2012. Investments, economics, psychology, politics, decision-making, human behavior, entrepreneurship and trend following are all passionately explored and debated on my show. To start? I'd like to give you a great piece of advice you can use in your life and trading journey… cut your losses! You will find much more about that philosophy here: https://www.trendfollowing.com/trend/ You can watch a free video here: https://www.trendfollowing.com/video/ Can't get enough of this episode? You can choose from my thousand plus episodes here: https://www.trendfollowing.com/podcast My social media platforms: Twitter: @covel Facebook: @trendfollowing LinkedIn: @covel Instagram: @mikecovel Hope you enjoy my never-ending podcast conversation!

Andrew Lo is author of “Adaptive Markets: Financial Evolution at the Speed of Thought.” He is also the Charles E. and Susan T. Harris Professor of Finance at MIT and the chairman and chief investment strategist of the AlphaSimplex Group. Andrew was taught from the beginning of his career that the efficient market hypothesis was gospel truth. It was the end-all-be-all. However, he first found a problem with the efficient market hypothesis just after graduating college. He did a test on the “random walk hypothesis” and related his findings from that hypothesis to the markets. He then came to find that his results proved the efficient market hypothesis wrong. Was there pushback during the early stages of talking about EMT being wrong? Absolutely. Andrew was one of the strongest that pushed back primarily because it went against everything he previously knew to be true. Andrew talks about another study he did with one of his MIT classes in 2004. He looked at hedge funds around that time and through data he proved that they were headed for trouble. They were able to foresee a small piece of the 2008 crash. Michael and Andrew end the podcast talking about Andrew’s new book and the role that the environment is playing in adaptive markets. When studying a species, what should be asked is, “Is it the species that is complex, or is it the environment that is complex and the species is just adapting to it?” Many species have figured out how to live in harsh environments in very different ways. In the same light, there are many different ways that people can trade the market and be successful. In this episode of Trend Following Radio: Efficient market hypothesis Adaptive markets hypothesis The random walk hypothesis Crowded trade phenomenon 2008 meltdown Paul Samuelson Commodities Corporation

Andrew Lo’s insights into how markets work lead him to tackle global issues. The MIT Sloan finance professor is using market levers to fix the problems of the world. Andrew discusses his new adaptive market theory, and how he’s using financial engineering to help cure cancer.

Bloomberg View columnist Barry Ritholtz interviews Andrew Lo, director of the Laboratory for Financial Engineering and the Charles E. and Susan T. Harris Professor at the MIT Sloan School of Management. Lo holds a bachelor's degree in economics from Yale University and a doctorate in economics from Harvard University. This commentary aired on Bloomberg Radio.

Cardiff Garcia talks to Alphaville's Matt Klein and FT senior investment commentator John Authers about the consequences and lessons of a famous call to sell stocks. Then, MIT economist Andrew Lo talks to John about the adaptive markets hypothesis, the subject of his forthcoming book. Clip courtesy of NBC. See acast.com/privacy for privacy and opt-out information.

Andrew Lo, Charles E. and Susan T. Harris Professor at MIT Sloan School of Management, and Director of MIT Laboratory for Financial Engineering, shares insight into the collaborative research efforts of MIT CSAIL and MIT Sloan School of Management within the three distinct areas of crytopgraphy, machine learning and AI, as well as discusses the progress of blockchain technology and cryptocurrency. He also offers a glimpse into the future of banking and finance and reveals the emerging technology of secured multi-party computation.

Andrew Lo straps sensors to traders and watches how their pulses and body temperatures change when markets dive or trades go bad. The technology could be used elsewhere in a bank to address problems before they escalate.

Click Here Or On Above Image To Reach Our ExpertsSecurity Expert Says, "We Can Now Spy On Human Emotions" Emotional surveillance has an undeniably dystopian vibe, like George Orwell's 1984, but it's not science fiction. Banks are already signing up for services that incorporate it into their analysis of behavior. A startup founded by MIT graduates called Humanyze has created a sensor-laden badge that transmits data on speech, activity, and stress patterns.One of these days, the walls may know when you're happy, sad, stressed or angry by using an experimental device unveiled Tuesday by researchers at the Massachusetts Institute of Technology that uses wireless signals to recognize emotions through subtle changes in breathing and heartbeat.Computer scientist Dina Katabi and her colleagues at the university's Computer Science and Artificial Intelligence Lab developed a radar system for vital signs that uses reflected radio signals to track movements, moods and behavior, with potential applications for smart homes, offices and hospitals.They posted their new research online Tuesday and plan to present their test results next month at a mobile-computing conference in New York.These wireless signals—a thousand times less powerful than conventional Wi-Fi—are designed to bounce off anyone within range, capturing variations in vital signs that can be analyzed quickly by a computer algorithm able to detect emotional states, the researchers said. To distinguish one mood from another, their system measures patterns of respiration, cardiac rhythms, and minute variations in the length of each individual heartbeat.“All of us share so much in how our emotions affect our vital signs,” said Dr. Katabi. “We get an accuracy that is so high that we can look at individual heartbeats at the order of milliseconds.”The system, which they call EQ-Radio, is 87% accurate at detecting whether a person is joyful, angry, sad or content, they said.By providing an accurate readout of moods, the system promises to loop people more directly into wireless sensor networks, the researchers said. While still experimental, the system could one day give buildings the capacity to respond automatically to changes in vital signs among the people living or working in them, without a need for explicit commands or a direct link to a body sensor, the researchers said.A hospital emergency room might automatically monitor patients awaiting treatment. An amusement park might modulate special effects by monitoring the involuntary reactions of people on an exhilarating ride. A house might one day react to a family's stress by playing pleasant music.PRO-DTECH II FREQUENCY DETECTOR(Buy/Rent/Layaway)“We have explored this idea of allowing a home to recognize someone's emotions and adapt to it,” said project researcher Fadel Adib. “The idea is to enable you to seamlessly interact with your home.”The team is already testing an earlier version of the system that tracks movements and behavior in about 15 homes in the Boston area, including that of Dr. Katabi. She uses it to monitor her sleep patterns and eating habits. It can track movements even if the person is in another room.“I would really like future homes to be more health aware,” she said.In the research made public Tuesday, Dr. Katabi and her colleagues tested the wireless system on 10 women and 20 men, between 19 and 77 years old, while in a standard office setting, which contained desks, chairs, couches and computers.CELLPHONE DETECTOR (PROFESSIONAL)(Buy/Rent/Layaway)During the tests, the volunteers sat from three to 10 feet away from the wireless sensors while attempting to evoke specific emotions by recalling emotion-rich memories. As a control, their vital signs during the experiment were also monitored using conventional electrocardiography and a video-based emotion recognition system that homes in on facial expressions.PRO-DTECH III FREQUENCY DETECTOR(Buy/Rent/Layaway)All told, the researchers collected measurements of 130,000 individual heartbeats. To classify the mood changes, the computer employed a machine learning algorithm to match the waveforms within each heartbeat.PRO-DTECH III FREQUENCY DETECTOR(Buy/Rent/Layaway)When they compared results, they found that the experimental system was almost as accurate in recognizing changes in emotion as the electrocardiograms. It was about twice as accurate as the facial cues recorded by the video system, they said.“We use the wireless signal to obtain the changes in the vital signs and then run a machine learning algorithm to get to emotions,” she said. “The algorithm can immediately recognize the emotions of someone new.”PRO-DTECH III FREQUENCY DETECTOR(Buy/Rent/Layaway)Wall Street Uses Technology To Spy On Traders Emotional StateThe trader was in deep trouble. A millennial who had only recently been allowed to set foot on a Wall Street floor, he made bad bets, and in a panic to recoup his losses, he'd blown through risk limits, losing $4.9 million in a single afternoon.WIRELESS/WIRED HIDDENCAMERA FINDER III(Buy/Rent/Layaway)It wasn't a career-ending day. The trader was taking part in a simulation run by Andrew Lo, an MIT finance professor. The goal: find out if top performers can be identified based on how they respond to market volatility. Lo had been invited into the New York-based global investment bank—he wouldn't say which one—after giving a talk to its executives. So in 2014, unknown to the outside world, he rigged a conference room with monitors to create a lab where 57 stock and bond traders lent their bodies to science.PRO-DTECH IV FREQUENCY DETECTOR(Buy/Rent/Layaway)Banks have already set up big-data teams to harvest insights from the terabytes of customer information they possess. Now they're looking inward to see whether they can improve operations and limit losses in their biggest cost center: employees. Companies including JPMorgan Chase and Bank of America have had discussions with tech companies about systems that monitor worker emotions to boost performance and compliance, according to executives at the banks who didn't want to be identified speaking about the matter.As machines encroach on humans' role in the markets, technology offers a way to even the fight. The devices Lo used—wristwatch sensors that measure pulse and perspiration—could warn traders to step away from their desks when their emotions run wild. They could also be used to screen hires to find those whose physiology is best suited to risk-taking—what interested the bank that allowed the MIT study.Wireless Camera Finder(Buy/Rent/Layaway)The most promising application, and the one with the most profound privacy issues, would be for keeping tabs on employees, Lo says. Risk managers could use it to spot problems brewing on a specific desk, such as unauthorized trading, before too much damage is done. “Imagine if all your traders were required to wear wristwatches that monitor their physiology, and you had a dashboard that tells you in real time who is freaking out,” Lo says. “The technology exists, as does the motivation—one bad trade can cost $100 million—but you're talking about a significant privacy intrusion.”MAGNETIC, ELECTRIC, RADIO ANDMICROWAVE DETECTOR(Buy/Rent/Layaway)Emotional surveillance has an undeniably dystopian vibe, like a finance version of George Orwell's 1984, but it's not science fiction. Banks are already signing up for services that incorporate it into their analysis of behavior. A startup founded by MIT graduates called Humanyze has created a sensor-laden badge that transmits data on speech, activity, and stress patterns.COUNTERSURVEILLANCE PROBE / MONITOR(Buy/Rent/Layaway)Microphones and proximity sensors on the gadgets help employers understand what high-performing teams are doing differently from laggards. The Boston-based company is close to announcing a deal with a bank that's moving some employees to new offices, according to Chief Executive Officer Ben Waber. The bank wants to use Humanyze badges to determine seating locations for traders, asset managers, and support staff to improve productivity, he says.Another startup, Behavox, uses machine-learning programs to scan employee communications and trading records. Emotional analysis of telephone conversations is a part of a worker's overall behavioral picture, according to founder Erkin Adylov, a former Goldman Sachs research analyst. When a worker deviates from established patterns—shouting at someone he's trading with when previous conversations were calm—it could be a sign further scrutiny is warranted. “Emotion recognition and mapping in phone calls is increasingly something that banks really want from us,” says Adylov, whose company is based in London. “All the things you do as a human are driven by emotions.”Emotions are reflexes that developed to drive behavior, scientists say, improving our prospects of seizing opportunity and surviving risk. They're accompanied by measurable physiological changes such as increased blood pressure, sweating, and a pounding heart. Their role in investing has been established since at least the time of economist Benjamin Graham, the father of value investing. More recently, John Coates, a University of Cambridge neuroscientist and former derivatives trader, has studied how financial risk takers' decisions are influenced by biology. His experiments, chronicled in a 2012 book, The Hour Between Dog and Wolf, show that hormones such as testosterone and cortisol play a part in exacerbating booms and busts.The volunteers in Lo's study were given a $3 million risk limit and told to make money in markets including oil, gold, stocks, currencies, and Treasuries. They came from across the bank's fixed-income and equity desks and ranged from junior employees to veterans with 15 years of experience. Top traders have a signature response to volatility, says Lo, who plans to publish his findings by next year. Rather than being devoid of feeling, they are emotional athletes. Their bodies swiftly respond to stressful situations and relax when calm returns, leaving them primed for the next challenge. The top performer made $1.1 million in a couple of hours of trading.Those who fared less well, like the trader who lost almost $5 million, were hounded by their mistakes and remained emotionally charged, as measured by their heart rate and other markers such as cortisol levels, even after the volatility subsided. Lo's findings suggest there's a sweet spot for emotional engagement: too much, and you're overly aggressive or fearful; too little, and you aren't involved enough to care. Veteran traders had more controlled responses, suggesting that training and experience count.There are other ways to infer emotional states. Researchers led by Kellogg School of Management professor Brian Uzzi pored over 1.2 million instant messages sent by day traders over a two-year period. They found that, as in Lo's study, having too much or too little emotion made for poor trades. Uzzi, whose study was published this year, says he's working with two hedge funds to design a product based on the research.As younger traders accustomed to biometric devices like the Fitbit enter the industry, applications designed to boost performance and monitor employees will become commonplace, says Lo, who expects it to be widespread in less than 10 years. “The more data we have, the more we're able to characterize the emotional state of the individual,” he says. “Everybody will have to have these kinds of analytics.”PRO-DTECH FREQUENCY DETECTOR(Buy/Rent/Layaway)Detecting Emotions In Thin AirOne of the most writerly things a person can do is to characterize air as thick, or emotions as tangible. Sadness lingers in the air. The best dinner parties are powered by palpable tension. The practice suggests that you are keenly attuned to your surroundings. Beyond observant, you use your senses in ways others had not thought possible. That is why people want to have sex with writers.But if you told me that the air is actually transmitting chemical signals that influence emotions between humans, I would add you to a list that I keep in my head. It's not a bad list, per se, but it is titled “Chumps.”One person who would not be on that list is Jonathan Williams. An atmospheric chemist, he describes himself as “one of those wandering scientific souls,” but not in an annoying way. He maintains a jovial British lilt after moving to Colorado to work at the National Oceanic and Atmospheric Administration, and then to Germany for a job with the Max Planck Institute (which describes itself as “Germany's most successful research organization”). There Williams and his colleagues study air.They focus on gases that come from vegetation in the tropics, as well as carbon industry. In doing so, the chemists use finely calibrated machines that sense the slightest changes in the contents of air. Taking measurements in the field, Williams and his colleagues always noticed that when they themselves got too close to the machines, everything went haywire.That made sense, in that humans are bags of gas. As breathing people know, we tend to emit carbon dioxide. (Though each exhalation still contains about four times as much oxygen as carbon dioxide.) And there are many subtler ingredients in the concoctions we breathe out. So Williams began to wonder, are these gases “significant on a global scale”? Could they be, even, contributing to climate change? Especially as the number of humans on Earth rockets toward 8 billion?The answer was no. Just a clear, simple no. By measuring gases in soccer stadiums, the Planck chemists found no consequence of human breath. There might be some effect at a global scale, but it's just nothing compared to the air-ravaging effects of transportation and agriculture.But Williams didn't come away from the stadium empty handed. As he sat and watched the fluctuating readings on the air sensors, he got an idea. In the manner of a typical European soccer crowd, the people went through fits of elation and anger,  joy and sorrow. So Williams began to wonder, as he later put it to me, “Do people emit gases as a function of their emotions?”If we do, it wouldn't be unprecedented. Tear some leaves off of a tree, for example, and it will emit chemical signals that may be part of a system of communication between trees. The behavior for bees and ants is clearly chemically dominated.“We're not like that—not like robots following chemicals,” Williams explained. “But it could be possible that we are influenced by chemicals emitted by other humans.”The idea of airborne pheromones—chemicals that specifically influence mating behaviors— has been a source of much fascination, but the actual evidence is weak. Some small studies have suggested an effect when people put cotton balls under their armpits, and then other people smell the balls—but in minor, unreliable ways.“I don't know why so many previous researchers have been so into armpits,” said Williams. “A much better way to communicate would be through your breath. Because you can direct your breath, and your breath is at roughly the same height as the person you're trying to communicate to, silently. In the dark, maybe, in your cave.” And if these behavior-modifying volatile chemicals exist (volatile meaning anything that goes into the air), then why would they be limited to sex? Why shouldn't we be able to signal fear or anxiety? It is true that birds seem to know that I'm afraid of them.Williams was so intrigued by the idea of gases and emotion that he designed another experiment—something more predictable than a German soccer game. This time he used a movie theater. Unlike the open-air stadium, the theater presented fewer variables. “You've got this box, the cinema, and you spool through air from outside at a continuous rate, and you have 250 people sitting there, not moving. And you show them all, simultaneously, something that should make them frightened or anxious or sad, or whatever.”The changes in any one person's breath might be minuscule, but a crowd of breathers could be enough to overcome the rest of the background signals. And more importantly, unlike a soccer match, the experiment could be done with the same film again and again. This could test the reproducibility of findings, which is critical to science.Rigging a mass spectrometer into the outflow vent of the theater, the Kino Cinestar in Mainz, Williams had a sense that the experiment as something of a lark. “I thought, we're probably just going to get a big mixture of popcorn and perfume,” he said. But, nonetheless, to measure relationships between scenes and gases, his team meticulously mapped out and labeled every scene in 16 films—from beginning to end. In 30 second increments, the team labeled each by its quality (kiss, pet, injury), as well as its emotional elements using a finite set of descriptors.

Our next guest is different than any guest we've had before, as she is not a fund manager but has spent much of her life's work researching and writing about the topic of trend following. She is a true thought leader in the managed futures industry and you'll learn a lot from the animated discussion we have regarding the history of trend following and how she co-authored her latest book on the subject.-----EXCEPTIONAL RESOURCE: Find Out How to Build a Safer & Better Performing Portfolio using this FREE NEW Portfolio Builder ToolIn This Episode, You'll Learn:About Kathryn's upbringing in Nashville, Tennessee.About her time at MIT where she went for electrical engineering.How her internship at a bank in France got her on the path to work in the financial industry.About her teaching financial engineering with Andrew Lo.Through her teaching and research, she became interested in technical analysis and trend following.How Kathryn went to the Stockholm School of Economics.When she left academia and joined RPM in Sweden.How she met Alex Greyserman and how she came to write a book with him.The history of trend following as she lays it out in her book.How trend following at its core is quite simple.What Kathryn likes to do outside of work.Her work life balance and her life in Sweden.Her view on the building blocks of trend following.Her quest for acceptance of trend following in the academic community.Where the term Crisis Alpha came from and what it is.About Convergent and Divergent strategies.About the Adaptive Market Hypothesis that she writes about in the book.What the CTA Smile is and what it really means.What she would look for when building or critiquing a research team.-----ATTENTION TTU TRIBE : SIGN-UP for Rick Rule's Symposium: Once in a life-time natural resource insights from the BEST investors in the world via a first-class livestream or Live event!Resources & Links Mentioned in this Episode:Kathryn's book is: Trend Following with Managed Futures.Learn more about Andrew Lo and MIT.Follow Niels on Twitter, LinkedIn, YouTube or via the TTU website.IT's TRUE ? – most CIO's read 50+ books each year – get your FREE copy of the Ultimate Guide to the Best Investment Books ever written here.And you can get a free copy of my latest book “The Many Flavors of Trend Following” here.Learn...

My guest today is Mikael Stenbom, the CEO and founder of RPM, which is an advisory consulting firm in Sweden. They advise, consult, or manage over 3.5B AUM. Stenbom has had tremendous experience in understanding strategy and the money management side of the CTA/managed futures world. The topic is CTA's. In this episode of Trend Following Radio we discuss: Stenbom's early influences Stenbom's economics background Connections between economics, trading, and sociology The Austrian school of economics His style of responsive, systematic trading to new clients Andrew Lo's Adaptive Markets Hypothesis The process of rebranding Stenbom's firm in 2008 The “politically infected” investor The cultural compatibility between Nordic countries and the Japanese Jump in! --- I'm MICHAEL COVEL, the host of TREND FOLLOWING RADIO, and I'm proud to have delivered 10+ million podcast listens since 2012. Investments, economics, psychology, politics, decision-making, human behavior, entrepreneurship and trend following are all passionately explored and debated on my show. To start? I'd like to give you a great piece of advice you can use in your life and trading journey… cut your losses! You will find much more about that philosophy here: https://www.trendfollowing.com/trend/ You can watch a free video here: https://www.trendfollowing.com/video/ Can't get enough of this episode? You can choose from my thousand plus episodes here: https://www.trendfollowing.com/podcast My social media platforms: Twitter: @covel Facebook: @trendfollowing LinkedIn: @covel Instagram: @mikecovel Hope you enjoy my never-ending podcast conversation!

Michael Covel speaks with Mikael Stenbom. Stenbom is the CEO and founder of RPM, which is a advisory consulting firm in Sweden. They advise, consult, or manage over 3.5B AUM. Stenbom has had tremendous experience in understanding strategy and the money management side of the CTA/managed futures world. Stenbom and Covel discuss what makes up a “smart money investor”; third party risk monitoring; and the life-cycle of CTA’s, hedge funds, and businesses in general. Nothing is constant in this world, and things change--Stenbom places a time axis on CTA’s to better manage for his clients. He also gives his opinions on why certain CTA’s have found such success. Covel and Stenbom discuss some of Stenbom’s early influences, such as the first time Stenbom had a “lightbulb moment” in the systematic world; Stenbom’s economics background; connections between economics, trading, and sociology; and the Austrian school of economics. Further topics include how Stenbom goes about explaining his style of responsive, systematic trading to new clients; Andrew Lo’s Adaptive Markets Hypothesis (which says that markets are for the most part efficient, but from time to time due to changes in the market ecology, become extremely inefficient); the process of rebranding Stenbom’s firm in 2008; the “politically infected” investor; and the cultural compatibility between Nordic countries and the Japanese. Free trend following DVD: www.trendfollowing.com/win.

Lessons from the financial crisis could lead to solutions far afield of financial markets. Andrew Lo will provide an overview of the crisis, describe the role mathematics played, and suggest how a deeper understanding of human nature may allow us to focus the power of global financial markets on key societal challenges.