Podcasts about Google Brain

Have you ever spent an afternoon wrestling with a Jupyter notebook, hoping that you ran the cells in just the right order, only to realize your outputs were completely out of sync? Today's guest has a fresh take on solving that exact problem. Akshay Agrawal is here to introduce Marimo, a reactive Python notebook that ensures your code and outputs always stay in lockstep. And that's just the start! We'll also dig into Akshay's background at Google Brain and Stanford, what it's like to work on the cutting edge of AI, and how Marimo is uniting the best of data science exploration and real software engineering. Episode sponsors Worth Search Talk Python Courses Links from the show Akshay Agrawal: akshayagrawal.com YouTube: youtube.com Source: github.com Docs: marimo.io Marimo: marimo.io Discord: marimo.io WASM playground: marimo.new Experimental generate notebooks with AI: marimo.app Pluto.jl: plutojl.org Observable JS: observablehq.com Watch this episode on YouTube: youtube.com Episode transcripts: talkpython.fm --- Stay in touch with us --- Subscribe to Talk Python on YouTube: youtube.com Talk Python on Bluesky: @talkpython.fm at bsky.app Talk Python on Mastodon: talkpython Michael on Bluesky: @mkennedy.codes at bsky.app Michael on Mastodon: mkennedy

How do you build a system for turning wild ideas into world-changing innovations? Astro Teller, Captain of Moonshots at X, The Moonshot Factory, has spent over 15 years leading Google's audacious innovation lab—the birthplace of Waymo, Google Brain, and other breakthrough projects.In this special episode, recorded live in Austin at SXSW, Astro shares the playbook to create a moonshot factory. You'll Learn:

Astro Teller is Alphabet’s Captain of Moonshots. He oversees projects at X – the moonshot factory behind innovations like Waymo and Google Brain. To celebrate X’s 15 years of pushing boundaries, Astro Teller decided to take listeners inside the factory. On The Moonshot Podcast, inventors and entrepreneurs behind breakthrough technologies reflect on their projects, both the highs and the lows. Teller sits down with Oz to discuss the process of experimentation, the importance of accepting failure and the future of innovation at Alphabet’s X.See omnystudio.com/listener for privacy information.

David is an OG in AI who has been at the forefront of many of the major breakthroughs of the past decade. His resume: VP of Engineering at OpenAI, a key contributor to Google Brain, co-founder of Adept, and now leading Amazon's SF AGI Lab. In this episode we focused on how far test-time compute gets us, the real implications of DeepSeek, what agents milestones he's looking for and more.[0:00] Intro[1:14] DeepSeek Reactions and Market Implications[2:44] AI Models and Efficiency[4:11] Challenges in Building AGI[7:58] Research Problems in AI Development[11:17] The Future of AI Agents[15:12] Engineering Challenges and Innovations[19:45] The Path to Reliable AI Agents[21:48] Defining AGI and Its Impact[22:47] Challenges and Gating Factors[24:05] Future Human-Computer Interaction[25:00] Specialized Models and Policy[25:58] Technical Challenges and Model Evaluation[28:36] Amazon's Role in AGI Development[30:33] Data Labeling and Team Building[36:37] Reflections on OpenAI[42:12] Quickfire With your co-hosts: @jacobeffron - Partner at Redpoint, Former PM Flatiron Health @patrickachase - Partner at Redpoint, Former ML Engineer LinkedIn @ericabrescia - Former COO Github, Founder Bitnami (acq'd by VMWare) @jordan_segall - Partner at Redpoint

Join JJ as he delves into AI Agents with the CEO of Lutra.ai, Jiquan Ngiam. Discover how their no-code AI platform is revolutionizing business automation, enhancing workflow efficiency, and empowering users with AI-driven superpowers. Learn about Jiquan's journey from Coursera and Google Brain to founding Lutra.ai, and get insights into the future of AI, self-driving cars, and the dynamic tech industry. Tune in for a live demo and explore how Lutra.ai makes complex tasks simpler and more streamlined. Available on all platforms: Apple, Spotify, and YouTube.Lutra:Learn more about Lutra.ai:https://lutra.ai/Grab thisLutra playbookFollow Jiquan:https://www.linkedin.com/in/jngiam/View JJ'sAI Agent Course: 00:00 Introduction to Lutra.ai00:27 Founder's Journey: From Coursera to Google Brain02:04 The Evolution of AI: Key Milestones09:20 The Future of AI and Self-Driving Cars13:20 Introducing Lutra.ai: Vision and Challenges19:59 Lutra.ai Demo: Automating Email Management24:31 Navigating AI Downtime24:43 AI in Document Processing25:25 Step-by-Step AI Task Management26:15 Advanced AI Capabilities27:12 Creating and Using Playbooks28:23 Integrations and Practical Applications30:33 Complex Workflows and Hierarchical Playbooks31:34 AI in Social Listening32:42 Challenges and Future of AI Adoption43:32 Tips for Professionals Embracing AI45:57 Conclusion and How to Get Started with LutraWant to learnhow to build AI Agents?

In this special guest episode of the Effortless Podcast, Amit Prakash sits down with Rajat Monga, the creator of TensorFlow and current Corporate Vice President of Engineering at Microsoft. With a career spanning Google Brain, founding Inference, and leading AI inferencing at Microsoft, Rajat offers a unique perspective on the evolution of AI. The conversation dives into TensorFlow's revolutionary impact, the challenges of building startups, the rise of PyTorch, the future of inferencing, and how transformative tools like GPT-4 and OpenAI's Gemini are reshaping the AI landscape.Key Topics and Chapter Markers:Introduction to Rajat Monga & TensorFlow Legacy [0:00]The inflection points in AI: TensorFlow's role and challenges [6:00]PyTorch vs. TensorFlow: A tale of shifting paradigms [16:00]The startup journey: Building Inference and lessons learned [27:00]Exploring O1 and advancements in reasoning frameworks [54:00]AI inference: Cost optimizations and hardware innovations [57:00]Agents, trust, and validation: AI in decision-making workflows [1:05:00]Rajat's personal journey: Tools for resilience and finding balance [1:20:00] Host:Amit Prakash: Co-founder and CTO at ThoughtSpot, formerly at Google AdSense and Bing, and a PhD in Computer Engineering. Amit has a strong track record in analytics, machine learning, and large-scale systems. Follow Amit on:LinkedIn - https://www.linkedin.com/in/amit-prakash-50719a2/  X (Twitter) - https://x.com/amitp42 Guest:Rajat Monga: He is a pioneer in the AI industry, best known as the co-creator of TensorFlow. He has held senior roles at Google Brain and Microsoft, shaping the foundational tools that power today's AI systems. Rajat also co-founded Inference, a startup focused on anomaly detection in data analytics. At Microsoft, he leads AI software engineering, advancing inferencing infrastructure for the next generation of AI applications. He holds a Btech Degree from IIT, Delhi. Follow Rajat on:LinkedIn - https://www.linkedin.com/in/rajatmonga/  X (Twitter) - https://twitter.com/rajatmonga Share Your Thoughts: Have questions or comments? Drop us a mail at EffortlessPodcastHQ@gmail.com Email: EffortlessPodcastHQ@gmail.com

Dr. Mike Schuster is the head of the AI Core team at Two Sigma, where he leads engineers and quantitative researchers in advancing AI technologies across the firm's investment strategies and internal efficiencies. With over 25 years of expertise in machine learning and deep learning, Mike has been at the forefront of AI trends in tech and finance. Prior to Two Sigma, he spent 12 years at Google, contributing to transformative projects like Google Translate as part of the Google Brain team. Dr. Schuster holds a PhD in Electrical Engineering from the Nara Institute of Science and Technology in Japan and is recognized as a pioneer whose work has significantly shaped the AI landscape.In this conversation, we discuss:The challenges and importance of building collaborative teams for complex AI systems in finance.Key differences between developing AI technologies in tech companies like Google versus finance firms like Two Sigma.The evolution of neural networks and their transformative impact on applications like Google Translate.The ethical considerations and risks of using AI in finance compared to other industries.Insights into data quality challenges and strategies for addressing bias in financial modeling.Predictions for the future of AI, focusing on efficiency, data quality, and practical advancements over the next five years.Resources:Subscribe to the AI & The Future of Work NewsletterConnect with Mike SchusterAI fun fact articleEpisode on how AI is diagnosing and treating sleep disorders  

Jiquan Ngiam, Co-Founder and CEO of Lutra AI, discusses his career journey Stanford University to eventually founding Lutra. He shares how Lutra helps streamline workflows by assisting with data prospecting, lead enrichment, and automating repetitive tasks. Jiquan also explores the balance between AI and human creativity in marketing, highlights his vision for making Lutra user-friendly for non-technical users, and encourages listeners to explore its potential for automating their workflows.   About Lutra AI Lutra aims to revolutionise automation and allow users to easily create AI-driven workflows. The platform simplifies complex processes, helping automate tasks and optimise work effortlessly. Whether you're managing data, streamlining operations, or integrating apps, Lutra makes automation accessible to everyone. Since its launch, Lutra has been empowering businesses to boost productivity and focus on what matters, eliminating the barriers of traditional workflow tools and delivering a seamless automation experience.   About Jiquan Ngiam Jiquan Niam is the CEO and Co-Founder of Lutra, an innovative automation platform. Before founding Lutra, Jiquan was a key contributor at Google Brain and studied at Stanford University where he achieved a PHD in Computer Science. Jiquan Niam is a driving force behind AI-driven automation and is passionate about making advanced technology accessible to all.   Time Stamps [00:00:18] - Jiquan provides some background to his career and why he founded Lutra. 00:02:44] - Overview of Lutra's Purpose and Functionality [00:09:36] - Enhancing Marketing Efforts with Timely Data [00:15:16] - User-Friendly Interface and Accessibility [00:20:23] - Marketing Strategy: Product-Led Growth Approach [00:23:27] - The Future of Marketing Roles with AI [00:26:19] - Advice for Young Marketers: Embrace Technology [00:28:30]- How to Get Started with Lutra   Quotes “I felt like education was this new superpower that I could give people.” Jiquan Ngiam, co-founder and CEO of Lutra "AI will not replace you, but a person who's using AI really well is going to do a lot more than you." Jiquan Ngiam, co-founder and CEO of Lutra "Help the team understand, investing into understanding this technology and using it... It's going to be potentially very game-changing." Jiquan Ngiam, co-founder and CEO of Lutra   Follow Jiquan: Jiquan Ngiam on LinkedIn: https://www.linkedin.com/in/jngiam/ Lutra AI website: https://lutra.ai/ Lutra AI on LinkedIn: https://www.linkedin.com/company/lutra-ai/   Follow Mike: Mike Maynard on LinkedIn: https://www.linkedin.com/in/mikemaynard/ Napier website: https://www.napierb2b.com/ Napier LinkedIn: https://www.linkedin.com/company/napier-partnership-limited/   If you enjoyed this episode, be sure to subscribe to our podcast for more discussions about the latest in Marketing B2B Tech and connect with us on social media to stay updated on upcoming episodes. We'd also appreciate it if you could leave us a review on your favourite podcast platform. Want more? Check out Napier's other podcast - The Marketing Automation Moment: https://podcasts.apple.com/ua/podcast/the-marketing-automation-moment-podcast/id1659211547

Startup Project Podcast: Building AI Agents for Knowledge Workers with Lutra AI Jiquan Ngiam joins Nataraj to discuss the future of AI, from the rise of deep learning to the potential of AI agents for knowledge workers. They delve into [Guest Name]'s experiences working with Andrew Ng at Coursera and Google Brain, where he witnessed the power of scaling up compute and data in pushing the boundaries of AI. Timestamps: * **0:00 - Introduction:** Nataraj welcomes [Guest Name] to the show and introduces his impressive background. * **2:28 - Working with Andrew Ng:** [Guest Name] shares his experience working with Andrew Ng, emphasizing Ng's foresight and focus on scaling up neural networks. * **6:15 - The Importance of Data and Compute:** [Guest Name] highlights how data and compute became key drivers in the success of AI, using the example of AlexNet's breakthrough in 2012. * **12:25 - Democratizing Education with Coursera:** [Guest Name] discusses the early days of Coursera and the team's vision for democratizing access to education, especially in fields like machine learning. * **17:55 - Google Brain and the Rise of Transformers:** [Guest Name] reflects on his time at Google Brain, where he witnessed the emergence of transformers and their potential for generalizing across modalities. * **21:24 - The Limits of Scaling:** [Guest Name] questions the future of AI scaling, suggesting that we may be approaching a point of diminishing returns due to data limitations and the difficulty of creating truly effective synthetic data. * **28:13 - The Need for Data on Physical Tasks:** [Guest Name] proposes a bold idea: collecting real-world data on mundane tasks to train AI agents for robotics and other applications that require replicating human behavior. * **34:23 - Lutrei.ai: AI Agents for Knowledge Work:** [Guest Name] introduces Lutrei.ai, an AI agent designed to assist knowledge workers with tasks like research, data manipulation, and automation. * **42:49 - Different Approaches to AI Agents:** [Guest Name] compares Lutrei's approach to building AI agents with other common methods, highlighting the importance of separating data and logic for reliable and scalable solutions. * **45:38 - Choosing the Right Models:** [Guest Name] discusses the diverse landscape of AI models and how Lutrei leverages different models for different tasks, from small models for summarization to larger models for reasoning and planning. * **52:04 - AI Code Generation: Cursor vs. GitHub Copilot:** [Guest Name] shares his experience using Cursor, a code generation tool, and compares it to GitHub Copilot, highlighting the potential for AI to empower average developers. * **1:00:16 - The Future of AI Code Generation:** [Guest Name] predicts that AI code generation capabilities will become ubiquitous, and the key innovations will be in user experience and interaction design. * **1:05:43 - Consuming Information:** [Guest Name] shares his favorite sources of information, including podcasts, books, and news outlets. * **1:08:44 - Mentorship and Learning:** [Guest Name] reflects on the key mentors in his career, including Andrew Ng, Daphne Koller, and John Chen. * **1:12:34 - Advice for Early Career Professionals:** [Guest Name] advises young professionals to be voracious learners and prioritize gaining diverse experiences early in their careers. * **1:16:21 - The Motivation Behind Lutrei:** [Guest Name] explains his passion for pushing the boundaries of AI while simultaneously making it accessible and impactful for a wider audience. * **1:18:33 - Closing Thoughts:** Nataraj thanks [Guest Name] for sharing his insights and expresses his excitement for the future of Lutrei.ai. **Don't miss this episode to learn more about the exciting things happening in gen AI and how it's poised to revolutionize the way we work!**

Before Elon Musk rebranded Twitter, X was already in use — at Google. Google X was Google's secret research lab, where Google's most imaginative ideas came to life. As CEO and co-founder, Astro Teller's job is to harness X's wildest, most futuristic technology to solve the world's hardest problems. The same "moonshot factory" that created Google Brain and Waymo self-driving cars is also working on carbon capture, laser-beam Internet, delivery drones, and more. I sat down with Astro to discuss how to build a culture of radical innovation. He shares some deep wisdom about unlearning what we know and why it's the counterintuitive approach that allows us to land a moonshot. This...is A Bit of Optimism. To learn more about Astro and his work, check out: X, the moonshot factorySee omnystudio.com/listener for privacy information.

Futurist, Technologist and Author of many titles including the classic “Wealth and Poverty”, George Gilder joins us to discuss supply side economics and the transformative potential of using graphene material in various industries including real estate. We discuss economic growth measured by time prices, showing that private sector progress is faster than GDP estimates. Learn about graphene's properties, including its strength and conductivity, and its potential to transform various industries. Graphene is a single layer of carbon atoms that is 200 times stronger than steel, 1000 times more conductive than copper and the world's thinnest material. Resources: getgilder.com Show Notes: GetRichEducation.com/517 For access to properties or free help with a GRE Investment Coach, start here: GREmarketplace.com Get mortgage loans for investment property: RidgeLendingGroup.com or call 855-74-RIDGE  or e-mail: info@RidgeLendingGroup.com Invest with Freedom Family Investments.  You get paid first: Text FAMILY to 66866 For advertising inquiries, visit: GetRichEducation.com/ad Will you please leave a review for the show? I'd be grateful. Search “how to leave an Apple Podcasts review”  GRE Free Investment Coaching: GREmarketplace.com/Coach Best Financial Education: GetRichEducation.com Get our wealth-building newsletter free— text ‘GRE' to 66866 Our YouTube Channel: www.youtube.com/c/GetRichEducation Follow us on Instagram: @getricheducation Complete episode transcript:   Automatically Transcribed With Otter.ai    Keith Weinhold  00:01 Welcome to GRE. I'm your host. Keith Weinhold. I'm talking about the various economic scare tactics out there, like the BRICS, the FDIC and the housing crash. What lower interest rates mean? How our nation's $35 trillion debt has gone galactic. Then today's guest is a legend. He's a technologist and futurist. It tells us about today's promise of graphene in real estate all today on get rich education.  when you want the best real estate and finance info, the modern Internet experience limits your free articles access, and it's replete with paywalls and you've got pop ups and push notifications and cookies disclaimers. Oh, at no other time in history has it been more vital to place nice, clean, free content in your hands that actually adds no hype value to your life. See, this is the golden age of quality newsletters, and I write every word of ours myself. It's got a dash of humor, and it's to the point to get the letter. It couldn't be more simple text, GRE to 66866, and when you start the free newsletter, you'll also get my one hour fast real estate course, completely free. It's called the Don't quit your Daydream letter, and it wires your mind for wealth. Make sure you read it. Text GRE to 66866, text GRE to 66866.   Corey Coates  01:40 you're listening to the show that has created more financial freedom than nearly any show in the world. This is Get Rich Education.   Keith Weinhold  01:56 Welcome to GRE from Dunedin, Florida to Dunedin, New Zealand and across 188 nations worldwide. I'm Keith Weinhold, and you are listening to get rich education, where real estate investing is our major. That's what we're here for, with minors in real estate economics and wealth mindset. You know, as a consumer of this media type as you are, it's remarkable how often you've probably encountered these de facto scare tactics, like the BRICS are uniting and it will take out the dollar and it's just going to be chaos in the United States. You might know that BRICS, B, R, I, C, S is the acronym for Brazil, Russia, India, China and South Africa. Do you know how hard it is to get off the petro dollar and how hard it is for the BRICS, which is basically more than just those five countries, it's dozens of countries. How hard it is for them to agree on anything with things as various as their different economies, and they'll have different customs and currencies. I mean, sheesh, just for you to get yourself and three friends all to agree to meet at the same coffee shop at the same time, takes, like a Herculean effort, plus a stroke of luck, and all full of you are like minded, so I wouldn't hold your breath on the dollar hyper inflating to worthlessness, although it should slowly debase. What about the scare tactic of the FDIC is going to implode, and this could lead to bank closures and widespread societal panic. Well, the FDIC, which stands for Federal Deposit Insurance Corporation, they're the body that backs all of the US bank deposits, including yours, and it's steered by their systemic resolution Advisory Committee. Well, there are $9 trillion in bank deposits, and is backed by only a few 100 billion in FDIC cash, so there aren't nearly enough dollars to back the deposits. So can you trust your money in the bank? That's a prevalence scare tactic, but my gosh, if nothing else, history has shown that the government will step in to backstop almost any crisis, especially a banking related one, where one failure can have a cascading effect and make other institutions fall. I'm not saying that this is right, but time has proven that the government does and will step in, or the common scare tactic in our core of the world that is the eminent housing price crash. And I define a crash as a loss in value of 20% or more. Do you know how difficult this would be to do anytime soon? Housing demand still outstrips supply. Today's homeowners have loads of protective equity, an all time high of about 300k so they're not walking away from their homes. Inflation has baked higher replacement costs into the real estate cake, and now mortgage rates have fallen one and a half percent from this cycle's highs, and they are poised to fall further, so a housing price crash is super unlikely, and a new scare tactic for media attention seems to be this proposal by a future presidential hopeful about a tax on unrealized gains. Now Tom wheelwright is the tax expert. He's returning to the show with us again soon here, so maybe I'll ask him about it. But a tax on unrealized gains is politically pretty unpopular. It would be a mess to impose, and a lot of others have proposed it in the past as well, and it has not gone anywhere. Plus tax changes need congressional approval, and we have a divided Congress, there's a small chance that attacks on unrealized gains could come to fruition, but it would be tough. It's probably in the category of just another media scare tactic, much like the BRICS and the shaky FDIC banking structure had a housing price crash. I like to keep you informed about these things, and at times we do have guests with a disparate opinion from mine on these things. Good to get a diversity of opinions, but it's best not to go too deep into these scare tactics that are really unlikely to happen any time soon. Well, there was a party going on 10 days ago at what all affectionately dub club fed in Jacksonhole Wyoming, I don't know what the club fed cover charge was, but fortunately, we did not have to watch Janet "Grandma" Yellen dance at Club fed and and share. Jerome Powell, yes, he finally caught a rate cut buzz. He announced that the time has come for interest rate cuts, and as usual, he didn't offer specifics. Total rager. what a party. later this month, he's going to render the long awaited decision, which now seems to be, how much will cut rates by a quarter point or a half point? Did you know that it's been four and a half years since the Fed lowered rates? Yeah, that was March of 2020, at the start of the pandemic. And then we know what happened back in 2022 and 2023 they hiked rates so much that they needed trail mix, a sleeping bag and some Mountain House freeze dried meals to go along with their steady hiking cycle. Interest rates now, though have been untouched for over a year, it's been an interesting year for the Fed and rates many erroneously thought there would be six or more rate cuts this year. And what about Maganomics? Trump recently said that if he becomes president, he should be able to weigh in on fed decisions that would depart from a long time tradition of Fed independence from executive influence. Historically, they've been separated.   Donald Trump  08:26 The Federal Reserve's a very interesting thing, and it's sort of gotten it wrong a lot. And he's tending to be a little bit later on things. He gets a little bit too early and a little bit too late. And, you know, that's very largely a it's a gut feeling. I believe it's really a gut feeling. And I used to have it out with him. I had it out with him a couple of times, very strongly. I fought him very hard. And, you know, we get along fine. We get along fine. But I feel that, I feel the president should have at least say in there. Yeah, I feel that strongly. I think that, in my case, I made a lot of money. Iwas very successful, and I think I have a better instinct than in many cases, people that would be on the Federal Reserve or the chairman.   Keith Weinhold  09:10 Those Trump remarks were just a few weeks ago, and then shortly afterward, he seemed to walk those comments back, but he did say that he would not reappoint. DJ J-pal, to the economic turntables. It's a long standing economic argument as well about whether an outside force like the Fed should set interest rates at all, which is the price of money, rather than allowing the rate to float with the free market as lenders and borrowers negotiate with each other. I mean, no one's out there setting the price of oil or refrigerators or grapes, but it is pretty remarkable that the Fed has signaled that rate cuts are eminent when inflation is still 2.9% well above their 2% target. But let's be mindful about the Fed's twofold mission, what they call their dual mandate. It is stable prices and maximum employment. Well, the Fed's concern is that second one, it's that the labor market has slowed and see the way it works is pretty simple. Lower interest rates boost employment because it's cheaper for businesses to borrow money that encourages them to expand and hire, which is exactly how lower interest rates help the labor market. That's how more people get hired, and this matters because you need a tenant that can pay the rent. So the bottom line here is to expect lower interest rates on savings accounts, HELOCs, credit cards and automobile loans. What this means to real estate investors is that lower mortgage rates are eminent, although the change should be slow. Two years ago, mortgage rates rose faster than they're going to fall. Now, one thing that lower interest rates can do is lower America's own debt. Servicing costs and America's public debt is drastic. Now, between 35 and $36 trillion in fact, to put our debt into perspective, it has gone galactic. And I mean that in an almost literal sense, because look, if you line up dollars, dollar bills, which are about six inches long, if you line those up end to end from Earth, how far do you think that they would reach? How about to the moon? Oh, no, if you line up dollars end to end, they would stretch beyond the moon. Okay, let's see how far we can follow them out through the solar system. They would breeze past Mars, which is 140 million miles away, the next planet out Jupiter. Oh, our trail of dollar bills would extend beyond that. Next up is Saturn and its ring. The dollar bills would reach beyond that. We're getting to the outer planets now, Uranus still going. Neptune, okay, Neptune is about $30 trillion bills away, and we would have to go beyond that then. So our 35 to $36 trillion of national debt would almost reach Pluto that's galactic. That's amazing. That's bad, and it probably means we have to print more dollars in order to pay back the debt, which is, of course, long term inflationary. And I don't know what's stopping us from going from $36 trillion up to say, 100 trillion, gosh. next week here on the show, we're talking about real estate investing in one of the long time best and still hottest real estate investor states, and then later on, we've got brilliant tax wizard Tom wheelwright returning, as we know here at GRE real estate pays five ways, and if you have any Spanish speaking family or friends, I've got a great way for them to consume all five video modules. It's an AI converting my voice to Spanish in these videos, we have a Spanish speaker here on staff at Get Rich Education, and she said the dub is pretty good. Well, the entire package, real estate pays five ways in Espanol is condensed into a powerful one hour total, all five videos a course, all in one wealth building hour. It's free to watch. There's no email address to enter or anything you can tell your Spanish speaking family and friends, or maybe your multilingual and your primary language is Spanish. That is it getricheducation.com/espanolricheducation.com/espanol or a shorter way to get to the same pageis getricheducation.com/espricheducation.com/esp, that's getricheducation.com/esp.richeducation.com/esp.  This week's guest is one of the first people I ever heard discussing the blockchain and cryptocurrency 15 years ago, and then he was early on AI. What got my attention is his education about a promising construction material for building new real estate, though, I expect that our discussion will delve outside of real estate today as well. Let's meet the incomparable George Gilder. This week's guest is the co founder at the Discovery Institute, discovery.org original pillar of supply side economics, former speechwriter to both Presidents Reagan and Nixon. And he's the author of the classic book on economics called Wealth and Poverty. Today he's at the forefront of technological breakthroughs. He's a Harvard grad. He wears a lot of stripes. I've only mentioned a few. Hey, welcome to GRE George Gilder.   George Gilder  15:09 right there better here.   Keith Weinhold  15:11 It's so good to host you, George, in both your writings and your influences on people like President Reagan, you champion supply side economics. And I think of supply side economics as things like lower taxes, less regulation and free trade. We had someone in the Reagan administration here with us a few months ago, David Stockman. He championed a lot of those same things. But go ahead and tell us more about supply side economics and what that means and how that's put into practice.   George Gilder  15:43 Well, it really begins with human creativity in the image of your Creator, essence of supply side economics now super abundant. I mean supply side economics triumphs. We had the whole information technology revolution ignited during the Reagan years and now dominates the world economy and gives the United States seven out of the top 10 companies in market cap. 70% of global corporate market cap is American companies because of supply side economics amazing, and that's why it's distressing to see supply side economics, with its promise of super abundance and prosperity and opportunity, Give way to narrow nationalistic calculations and four tenths of war. I mean, all these Jews are at the forefront. Today, in time, we're going to see human creativity once again prevail in my books, Life After Capitalism is my latest book, my new paradigm is graphene. Graphene is a single layer of carbon atoms, two dimensional layer of carbon atoms that is 200 times stronger than steel, 1000 times more conductive than copper. It switches and the terahertz trillions of times a second, rather than the billions of times a second that our current silicon chips which and you mix it with concrete, the concrete comes 35% stronger, just parts per million of graphene mixed with concrete yields some material that's 35% stronger than ordinary concrete. You mix a parts per million of graphene with asphalt, the roads don't get potholes in the winter. It's radically Abate, but it conducts signals so accurately. If you go on YouTube, you can find a mouse and said it's spinal cord severed completely, injected with graphene, the spinal signals transmitted so accurately that the you see the mouse doing cartwheels by the end of the YouTube measure. I mean, it's material that's going to transform all industries, from real estate to medicine to surgery to electronics. Electronics been kind of the spearhead of our economy, of the transformation and electronics may be more significant than any other domain.   Keith Weinhold  18:49 Well, this is a terrific overview of all the contributions you're making to both the economic world and the technology world with what you told us about right there. And I do want to ask you some more about the graphene and the technology later. But you know, if we bring it back to the economics, it was in your classic book, Wealth and Poverty, which sold over a million copies, where you espouse a lot of the same things that you still espouse today in your more recent books, that is, capitalism begins with giving, we can often think of it that way. As a real estate investor is where we need to give tenants a clean, safe, affordable, functional property before we profit. Capitalism begins with giving.   George Gilder  19:32 Absolutely. That's a crucial debate I had with Ayn Rand The Fountainhead and Atlas Shrugged and I say, capitalism is subsist on altruism. I'm concerned for the interests of others, imaginative anticipation of the needs of others. It's an altruistic, generous system, and from that generosity. Stems the amazing manifestations of super abundance that which I've been writing about recently. And super abundance shows, measured by time prices, how many hours a typical worker has to spend to earn the goods and services that sustain its life. Yeah, that's where the real cost has time. Yeah, time is money. Money is time, tokenized time, and measured by time, economic growth has been 50 to just enormously faster than is estimated by any of the GDP numbers. However, measured by time government services or ordinary GDP assumes that every dollar of government spending is worth what it costs. Prices both show that progress in the private sector has been four or five times faster than is estimated by GDP well government time, price of government dominated goods, including, increasingly, healthcare and education, is way less valuable than the cost. It's value subtracted, and certainly trillions of dollars for windmills and solar panels, trillions of dollars of subsidies is a net subtraction of value in the world economy. So I am with Gale Pooley and Tupy, both who wrote a book called Superabundance that I wrote the introduction to, and William Nordhaus, the Nobel laureate from Yale, who really conceived and developed time prices and showed that economic growth is 1000s of times greater than has been estimated by ordinary economic data. This is a time of abundance. It's not a time of scarcity. It's not a time of the dismal science. It's the time of super abundance.   Keith Weinhold  22:17 Yes, 100% a lot of that is just the government getting out of the way and really let people be givers, be that go giver and lead with giving, because I have never heard of a society that's taxed its way to prosperity.   George Gilder  22:34 Yeah. Well, that's absolutely the case. And I've been talking previously about graphene, which is the great new material that has been discovered of the last a couple decades. It originated, a lot of the science originated in Jim Tour's laboratory. James Tour of Rice University, and he's had scores of companies have emerged from his laboratory, and 18 of them got started in Israel. Israel is really become a leading force in the world economy. And when Israel is in jeopardy, our economy is in jeopardy. We have 100,000 Israeli citizens working in companies in Silicon Valley, 100,000 all the leading American tech companies have outposts in Israel, and now we face what I call the Israel test, which is how you respond to people who are really superior in creativity and accomplishment and intellect, and the appropriate thing to do is emulate them and learn from them. But too many people in the world see success and they want to tear it down, or they think it was stolen from someone else, or it was part of a zero sum game where the riches of one person necessarily come at the expense of someone else, which is the opposite of the truth, the riches proliferate opportunities for others. That's how the economy grows through the creativity and the image of your Creator.   Keith Weinhold  24:25 And when you bring up Israel, they're one of many nations that's made strong contributions to society and the economy, and we think about other nations that's been an increasingly relevant conversation these past few years, a lot of that centers on immigration. I'm not an expert on how many people we should let into this country or any of those sort of policy sorts of things, but here is a real estate investing show. I often think about where and how we're going to house all these immigrants, whether they come from Central America or South America or Israel or. Anywhere else. And I know oftentimes you've touted immigrations economic benefits, so I think it's pretty easy for one to see how in the short term, immigrants could be of economic detriment, but tell us more about those long term economic benefits of immigrants coming to the United States.   George Gilder  25:17 Immigrants come to the United States and become Americans and contribute American opportunity and wealth. We won the second world war because of immigration of Jewish scientists from Europe to the United States, who led by people like John von Neumann and Oppenheimer who forged the Manhattan Project, and that's really how we won the Second World War, was by accepting brilliant immigrants who wanted to serve America. Now there is a threat today where immigrants come to the United States not to contribute to the United States, but to exploit the United States, or even destroy it, not to go givers. They are givers, and so we want immigrants who are inclined to commit to America and create opportunities for the world, but immigrants who want to tear down America and who believe that America owes them something tend to be less productive and less valuable immigrants and immigrants who really want to destroy western civilization, and the jihadists that we know about are actually a threat to America. So the immigration problem isn't simple, but when we had a system where legal immigrants could apply and enter our country and revitalize it, that was a wonderful system, but having boards of illegal immigrants just pour over the border is not an intelligent way to deal with the desire of people around the world to share an American prosperity.   Keith Weinhold  27:13 We've seen several cases in the past year or two where immigrants are given free housing. There are really great case studies about this in Massachusetts and some other places, how they're giving housing before oftentimes, our own Americans, including sometimes retired veterans, are provided with housing. This all comes down to the housing crunch and already having a low housing supply. So what are some more your thoughts about just how much of a layup or a handout should we give new immigrants?   George Gilder  27:42 Housing technology is going to be transformed by the material science revolution that is epitomized by graphene, this miracle material I was describing. I think part of the problem is real estate enterprise is over regulated, and there are too many obstacles to the building of innovative new forms of housing. In 20 years, it'll be hard to recognize many of the structures that emerge as a result of real revolution in material science that is epitomized by this graphene age that I've been describing, and that also will transform electronics as well, and part housing can become a kind of computer platform as Elon Musk is transforming the auto business by seeing Tesla is really a new form of computer platform. I believe there's going to be an Elon Musk of real estate who is going to re envisage housing as a new form of building a computer platform that makes intelligent houses of the future that will be both cheaper and more commodious for human life.   Keith Weinhold  29:12 Real estate is rather old and slow moving when we think about technology in real estate, maybe what comes to mind are smart thermostats, smart doorbells, or 3d printed homes. When we come back, we're going to learn more about graphene and what it can do in real estate in the nanocosm revolution. Our guest is George Gilder. We talked about economics. We're coming back to talk about technology. I'm your host. Keith Weinhold. Keith Weinhold  Hey, you can get your mortgage loans at the same place where I get mine, at Ridge lending group NMLS, 42056, they provided our listeners with more loans than any provider in the entire nation because they specialize in income properties. They help you build a long term plan for growing your real estate empire with less. Ridge you can start your pre qualification and chat with President Caeli Ridge personally. Start now while it's on your mind at ridgelendinggroup.com That's ridgelendinggroup.com. Your bank is getting rich off of you. The national average bank account pays less than 1% on your savings. If your money isn't making 4% you're losing your hard earned cash to inflation. Let the liquidity fund help you put your money to work with minimum risk, your cash generates up to an 8% return with compound interest year in and year out, instead of earning less than 1% sitting in your bank account, the minimum investment is just 25k you keep getting paid until you decide you want your money back. Their decade plus track record proves they've always paid their investors 100% in full and on time. And I would know, because I'm an investor too, earn 8% hundreds of others are text FAMILY to 66866, learn more about freedom. Family Investments Liquidity Fund on your journey to financial freedom through passive income. Text FAMILY to 66866.   Dolf Deroos  31:19 This is the king of commercial real estate. Dolph de Roos, listen to get rich education with Keith Weinhold, and don't quit your Daydream.   Keith Weinhold  31:32 Welcome back to Get Rich Education. We're joined by an illustrious, legendary guest, George Gilder, among being other things, including a prolific writer. He's also the former speechwriter to presidents Reagan and Nixon. He's got a really illustrious and influential career. George, you've been talking about graphene, something that I don't think our audience is very familiar with, and I'm not either. Tell us about graphene promise in real estate.   George Gilder  31:59 Well, back in Manchester, England, in 2004 graphene was first discovered and formulated. It actually was submerged before then, but the Nobel Prizes were awarded to Geim and Novoselov in2010.  So this is a new material that all of us know when we use a lead pencil, a lead is graphite, and graphene is a single layer of graphite. And it turns out, many people imagined if you had a single layer of graphite, it would just break up. It would not be useful.   Keith Weinhold  32:42 We're talking super thin, like an atom.   George Gilder  32:45 Yeah, it's an atom thick, but still, it turns out that it has miraculous properties, that it's 200 times stronger than steel. If you put it in a trampoline, you couldn't see the trampoline, but you could bounce on it without go following through it. It can stop bullets. It means you can have invisible and almost impalpable bulletproof vests, and you mix it with concrete, and the concrete is becomes 35% stronger, even parts per million of graphene can transform the tensile strength of concrete, greatly reduce the amount you need, and enable all sorts of new architectural shapes and capabilities. We really are in the beginning of a new technological age, and all depressionary talk you hear is really going to be eclipsed over coming decades by the emergence of whole an array of new technologies, graphene, for instance, as a perfect film on wafer of silicon carbide and enable what's called terahertz electronics, which is trillions of cycles a second like light rather than billions of cycles a second like or Nvidia or L silicon chips, and it really obviates chips, because you what it allows is what's called wafer scale integration of electronics, and today, it the semiconductor industry, and I've written 10 books on semiconductors over the years, but the semiconductor industry functions by 12 inch wafers that get inscribed with all sorts of complex patterns that are a billionth of a meter in diameter. These big wafers and then the way. First get cut up into 1000s of little pieces that each one gets encapsulated in plastic packages and by some remote Asian islands, and then get implanted on printed circuit boards that arrayed in giant data centers that now can on track to consume half the world's energy over the next 20 years, and these new and all this technology is ultimately going to be displaced by wafer scale integration on The wafer itself. You can have a whole data center on a 12 inch wafer with no chips. It's on the wafer itself. And this has been recently announced in a paper from Georgia Tech by a great scientist named Walter de Heere. And it's thrilling revolution that that render as much as Silicon Valley obsolescent and opens up just huge opportunities in in construction and real estate and architecture and medicine and virtually across the range of contemporary industry.   Keith Weinhold  36:20 You wrote a book about blockchain and how we're moving into the post Google world is what you've called it. So is this graphene technology that you're discussing with us here? Is that part of the next thing, which you're calling the nanocosm revolution?   George Gilder  36:36 The microcosm was an earlier book the quantum revolution and economics and technology. I thought I wrote years ago called microcosm.   Keith Weinhold  36:46 Okay, we're getting smaller than microcosm now in nanocosm.   36:49 that was microns, that was millionths of a meter dimensions of the transistors and devices and silicon chips, the nanocosm is a billionth of the meter. It's 1000 times smaller the features and electronics of the future, and we're moving from the microcosm into the nanocosm. New materials like graphene epitomize this transformation. You know, people think that these giant data centers all around the world, which are amazing structures, but half the energy in these data centers are devoted to removing the heat rather than fueling the computation. And I believe these data centers are represent a kind of IBM mainframe of the current era. When I was coming up, people imagined that a few 100 IBM mainframe computers, each weighing about a ton, would satisfy all the world's needs for computation, and that new artificial minds could be created with these new IBM mainframes. And it's the same thing today, only we're talking about data centers, and I believe that the coming era will allow data centers in your pocket and based on graphene electronics, and wait for scale integration, a whole new paradigm that will make the current data centers look like obsolete, old structures that need to be revitalized.   Keith Weinhold  38:37 Around 2007 Americans and much of the world, they got used to how it feels to have the power of a computer in their pocket with devices like the iPhone. How would it change one's everyday life to have effectively a data center in their pocket?   38:54 This means that we no longer would be governments of a few giant companies hearing a singular model of intelligence. That's what's currently envisaged, that Google Brain or Facebook or these giant data setters would sum up all human intelligence and in a particular definition, but there are now 8 billion human beings on earth, and each of our minds is as densely connected as the entire global internet. And while the global Internet consumes error watts, trillions of watts of power, or brains. Each of these 8 billion human minds functions on 12 to 14 watts, or it's billions of times less than these data center systems. On the internet. I believe that technology works to the extent that it expands human capabilities, not to the extent that it displaces human capabilities. The emergence of distributed databases in all our pockets, distributed knowledge and distributed creativity can revitalize the whole world economy and open new horizons that are hard to imagine today, as long as we don't, all of a sudden decide that we live in a material universe where everything is scarce and successes by one person come at the expense of somebody else, as long as that zero sum model doesn't prevail, right? Human opportunities are really unlimited. Most of economics has been based on a false model of scarcity, the only thing that's really scarce is time. Imagination and creativity are really infinite.   Keith Weinhold  41:10 Yes, well, if someone wants to learn more about graphene in the nanocosm revolution, how can you help them? What should they do?   41:18 They can read my newsletters. I have a company with four newsletters. I write the Gilder Technology Report. Much of the time I write, John Schroeder writes moonshots, which is and I have a Gilder Private Reserve that reaches out with our crowd and Israel, and a lot of those graph gene companies in Israel are part of our Private Reserve. And I do Gilders Guide posts, and those are all available getgilder.com.   Keith Weinhold  41:56 if you'd like to learn more about George and his popular newsletter called the Gilder Technology Report. You can learn more about that at get gilder.com George, it's been an enlightening conversation about economics and where society is moving next. Thanks so much for coming on to the show.   George Gilder  42:16 Thank you, Keith. I really appreciate it.   Keith Weinhold  42:24 yeah, a forward looking discussion with the great George Gilder. Forbes said graphene may be the next multi trillion dollar material. George will tell you that you want to get into graphene now, while the biggest gains are still ahead. If it interests you in at least learning more, check out his video resource. It's free. There's also an opportunity for you to be an investor. You can do all of that and more at getgilder.com again getguilder.com until next week. I'm your host. Keith Weinhold. Don't Quit Your Daydream.   43:04 nothing on this show should be considered specific, personal or professional advice. Please consult an appropriate tax, legal, real estate, financial or business professional for individualized advice. Opinions of guests are their own. Information is not guaranteed. All investment strategies have the potential for profit or loss. The host is operating on behalf of Get Rich Education LLC, exclusively.   Keith Weinhold  43:32 The preceding program was brought to you by your home for wealth building. GetRichEducation.com

Aidan Gomez is the Co-founder & CEO at Cohere, the leading AI platform for enterprise, having raised over $1BN from some of the best with their last round pricing the company at a whopping $5.5BN. Prior to Cohere, Aidan co-authored the paper “Attention is All You Need,” which introduced the groundbreaking Transformer architecture. He also collaborated with a number of AI luminaries, including Geoffrey Hinton and Jeff Dean, during his time at Google Brain, where the team focused their efforts on large-scale machine learning. In Today's Episode with Aidan Gomez We Discuss: 1. Compute vs Data: What is the Bottleneck: Does Aidan believe that more compute will result in an equal increase in performance?   How much longer do we have before it becomes a case of diminishing returns?   What does Aidan mean when he says "he has changed his mind massively on the role of data"? What did he believe? How has it changed? 2. The Value of the Model: Given the demand for chips, the consumer need for applications, how does Aidan think about the inherent value of models today? Will any value accrue at the model layer? How does Aidan analyze the price dumping that OpenAI are doing? Is it a race to the bottom on price? Why does Aidan believe that "there is no value in last year's model"? Given all of this, is it possible to be an independent model provider without being owned by an incumbent who has a cloud business that acts as a cash cow for the model business? 3. Enterprise AI: It is Changing So Fast: What are the biggest concerns for the world's largest enterprises on adopting AI? Are we still in the experimental budget phase for enterprises? What is causing them to move from experimental budget to core budget today? Are we going to see a mass transition back from Cloud to On Prem with the largest enterprises not willing to let independent companies train with their data in the cloud? What does AI not do today that will be a gamechanger for the enterprise in 3-5 years? 4. The Wider World: Remote Work, Downfall of Europe and Relationships: Given humans spending more and more time talking to models, how does Aidan reflect on the idea of his children spending more time with models than people? Does he want that world? Why does Aidan believe that Europe is challenged immensely? How does the UK differ to Europe? Why does Aidan believe that remote work is just not nearly as productive as in person?  

On this episode of FYI, ARK's Chief Futurist Brett Winton, and Chief Investment Strategist Charlie Roberts sit down with artificial intelligence (Al) luminary Andrew Ng to explore the deployment of artificial intelligence and the evolution of AI education. Andrew shares insights from his extensive career, including his work with Google Brain, Baidu, Coursera, and his current AI fund. We analyze the transformative potential of AI, especially in how large corporations can harness it, the progression toward agentic systems, and the contentious topic of open-source AI. This episode provides a comprehensive overview of AI's current status and future trajectory, offering invaluable insights for technology enthusiasts."For the last 10-15 years, there have constantly been a small number of voices saying AI is hitting a wall. I think that a lot of statements to that effect were all over and over proven to be wrong. I think we're so far from hitting a wall." -Andrew Ng Key Points From This Episode:- Andrew Ng's significant contributions to AI and education through various platforms- Insights into the deployment challenges and future potentials of AI in business- The role of agentic systems in advancing AI applications- The impact of open source on innovation and the AI industry- Distribution and data generation in AI's effectiveness

Hugo speaks with Shreya Shankar, a researcher at UC Berkeley focusing on data management systems with a human-centered approach. Shreya's work is at the cutting edge of human-computer interaction (HCI) and AI, particularly in the realm of large language models (LLMs). Her impressive background includes being the first ML engineer at Viaduct, doing research engineering at Google Brain, and software engineering at Facebook. In this episode, we dive deep into the world of LLMs and the critical challenges of building reliable AI pipelines. We'll explore: The fascinating journey from classic machine learning to the current LLM revolution Why Shreya believes most ML problems are actually data management issues The concept of "data flywheels" for LLM applications and how to implement them The intriguing world of evaluating AI systems - who validates the validators? Shreya's work on SPADE and EvalGen, innovative tools for synthesizing data quality assertions and aligning LLM evaluations with human preferences The importance of human-in-the-loop processes in AI development The future of low-code and no-code tools in the AI landscape We'll also touch on the potential pitfalls of over-relying on LLMs, the concept of "Habsburg AI," and how to avoid disappearing up our own proverbial arseholes in the world of recursive AI processes. Whether you're a seasoned AI practitioner, a curious data scientist, or someone interested in the human side of AI development, this conversation offers valuable insights into building more robust, reliable, and human-centered AI systems. LINKS The livestream on YouTube (https://youtube.com/live/hKV6xSJZkB0?feature=share) Shreya's website (https://www.sh-reya.com/) Shreya on Twitter (https://x.com/sh_reya) Data Flywheels for LLM Applications (https://www.sh-reya.com/blog/ai-engineering-flywheel/) SPADE: Synthesizing Data Quality Assertions for Large Language Model Pipelines (https://arxiv.org/abs/2401.03038) What We've Learned From A Year of Building with LLMs (https://applied-llms.org/) Who Validates the Validators? Aligning LLM-Assisted Evaluation of LLM Outputs with Human Preferences (https://arxiv.org/abs/2404.12272) Operationalizing Machine Learning: An Interview Study (https://arxiv.org/abs/2209.09125) Vanishing Gradients on Twitter (https://twitter.com/vanishingdata) Hugo on Twitter (https://twitter.com/hugobowne) In the podcast, Hugo also mentioned that this was the 5th time he and Shreya chatted publicly. which is wild! If you want to dive deep into Shreya's work and related topics through their chats, you can check them all out here: Outerbounds' Fireside Chat: Operationalizing ML -- Patterns and Pain Points from MLOps Practitioners (https://www.youtube.com/watch?v=7zB6ESFto_U) The Past, Present, and Future of Generative AI (https://youtu.be/q0A9CdGWXqc?si=XmaUnQmZiXL2eagS) LLMs, OpenAI Dev Day, and the Existential Crisis for Machine Learning Engineering (https://www.youtube.com/live/MTJHvgJtynU?si=Ncjqn5YuFBemvOJ0) Lessons from a Year of Building with LLMs (https://youtube.com/live/c0gcsprsFig?feature=share) Check out and subcribe to our lu.ma calendar (https://lu.ma/calendar/cal-8ImWFDQ3IEIxNWk) for upcoming livestreams!

Crucible Moments will be back shortly with season 2. You'll hear from the founders of YouTube, DoorDash, Reddit, and more. In the meantime, we'd love to introduce you to a new original podcast, Training Data, where Sequoia partners learn from builders, researchers and founders who are defining the technology wave of the future: AI. The following conversation with Harrison Chase of LangChain is all about the future of AI agents—why they're suddenly seeing a step change in performance, and why they're key to the promise of AI.  Follow Training Data wherever you listen to podcasts, and keep an eye out for Season 2 of Crucible Moments, coming soon. LangChain's Harrison Chase on Building the Orchestration Layer for AI Agents Hosted by: Sonya Huang and Pat Grady, Sequoia Capital Mentioned:  ReAct: Synergizing Reasoning and Acting in Language Models, the first cognitive architecture for agents SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering, small-model open-source software engineering agent from researchers at Princeton Devin, autonomous software engineering from Cognition V0: Generative UI agent from Vercel GPT Researcher, a research agent  Language Model Cascades: 2022 paper by Google Brain and now OpenAI researcher David Dohan that was influential for Harrison in developing LangChain Transcript: https://www.sequoiacap.com/podcast/training-data-harrison-chase/

2024: The Most Important Year in the History of Robotics!Companion podcast #31 to Keynote address at SuperTechFT 3 July 2024 Happy to be with you one and all. I'm Tom Green, your host and companion on this very special journey for 2024. We are only halfway through the year, and already 2024 has shown us that it is the most important year in the history of robotics.This podcast will show you why that is.This podcast is a companion to the live keynote address I will give at SuperTechFT in San Francisco on July 3rd 2024. I want to first thank Dr. Albert Hu, president and director of education at SuperTechFT, and to the staff and patrons of SuperTechFT for inviting me. The title of my keynote: 2024: The Most Important Year in the History of Robotics!What other year can possibly compete for top honors other than 2024?2024 eliminated the barrier to entry for digital programming by eliminating the need to code.As Tesla's former chief of AI, Andrej Karpathy put it: "Welcome to the hottest new programming language...English"2024 opened the door of AI prompt engineering to millions of new jobs and careers in millions of SME industries worldwide.So explains: Andrew Ng, investor and former head of Google Brain and Baidu.2024 converged GenAI with robotics, broadened robot/cobot applications, and freed robots from complexity of operation.So announced NVIDIA's CEO and founder Jensen Huang at the company's March meeting.2024 reinvigorated the liberal arts, creative thinking, expository writing, and language as vital new components in developing robotics applications.So reflects Stephen Wolfram physicist and creator of Mathematica2024 defined the need for the GenAI & the "New Collar" Worker Connection: Vitally needed workers for AI/robot-driven industry worldwide, and just maybe, the revitalization of America's middle class…or the middle class of any nation.Sarah Boisvert technologist, factory owner and wrote the book on the New Collar WorkforceSuddenly in mid-2024, technology has thrown us into a brand-new worldAnd it's only early July of 2024...can you believe it?“Artificial intelligence and robotics could catapult both fields to new heights.”The 4-Year Plight: SMEs in Search of Robots!Tech News May Fade, but Its Stories Are Forever! GenAI & "New Collar" ConnectionDid AI Just Free Humanity from Code?

David Luan is the CEO and Co-Founder at Adept, a company building AI agents for knowledge workers. To date, David has raised over $400M for the company from Greylock, Andrej Karpathy, Scott Belsky, Nvidia, ServiceNow and WorkDay. Previously, he was VP of Engineering at OpenAI, overseeing research on language, supercomputing, RL, safety, and policy and where his teams shipped GPT, CLIP, and DALL-E. He led Google's giant model efforts as a co-lead of Google Brain. In Today's Episode with David Luan We Discuss: 1. The Biggest Lessons from OpenAI and Google Brain: What did OpenAI realise that no one else did that allowed them to steal the show with ChatGPT? Why did it take 6 years post the introduction of transformers for ChatGPT to be released? What are 1-2 of David's biggest lessons from his time leading teams at OpenAI and Google Brain? 2. Foundation Models: The Hard Truths: Why does David strongly disagree that the performance of foundation models is at a stage of diminishing returns? Why does David believe there will only be 5-7 foundation model providers? What will separate those who win vs those who do not? Does David believe we are seeing the commoditization of foundation models? How and when will we solve core problems of both reasoning and memory for foundation models? 3. Bunding vs Unbundling: Why Chips Are Coming for Models: Why does David believe that Jensen and Nvidia have to move into the model layer to sustain their competitive advantage? Why does David believe that the largest model providers have to make their own chips to make their business model sustainable? What does David believe is the future of the chip and infrastructure layer? 4. The Application Layer: Why Everyone Will Have an Agent: What is the difference between traditional RPA vs agents? Why is agents a 1,000x larger business than RPA? In a world where everyone has an agent, what does the future of work look like? Why does David disagree with the notion of "selling the work" and not the tool? What is the business model for the next generation of application layer AI companies?  

Reed Albergotti is the tech editor at Semafor. He joins Big Technology Podcast to break down the week's news. We cover: 1) NVIDIA temporarily becoming the most valuable publicly traded company 2) Is NVIDIA a bubble? 3) What might disrupt NVIDIA? 4) Ilya Sustkever founds Safe Superintelligence Inc. 5) Who's funding Ilya? 6) Will SSI amount to anything? 6) OpenAI might become a public benefit company 7) The state of DeepMind's merger with Google Brain 8) Mustafa Suleyman's entry to Microsoft and how it impacts the relationship with OpenAI 9) Apple Vision Pro hits a speed bump 10) Vision Pro & Apple Intelligence similarities and differences ---- You can subscribe to Big Technology Premium for 25% off at https://bit.ly/bigtechnology Enjoying Big Technology Podcast? Please rate us five stars ⭐⭐⭐⭐⭐ in your podcast app of choice. For weekly updates on the show, sign up for the pod newsletter on LinkedIn: https://www.linkedin.com/newsletters/6901970121829801984/ Questions? Feedback? Write to: bigtechnologypodcast@gmail.com

Last year, AutoGPT and Baby AGI captured our imaginations—agents quickly became the buzzword of the day…and then things went quiet. AutoGPT and Baby AGI may have marked a peak in the hype cycle, but this year has seen a wave of agentic breakouts on the product side, from Klarna's customer support AI to Cognition's Devin, etc. Harrison Chase of LangChain is focused on enabling the orchestration layer for agents. In this conversation, he explains what's changed that's allowing agents to improve performance and find traction.  Harrison shares what he's optimistic about, where he sees promise for agents vs. what he thinks will be trained into models themselves, and discusses novel kinds of UX that he imagines might transform how we experience agents in the future.      Hosted by: Sonya Huang and Pat Grady, Sequoia Capital Mentioned:  ReAct: Synergizing Reasoning and Acting in Language Models, the first cognitive architecture for agents SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering, small-model open-source software engineering agent from researchers at Princeton Devin, autonomous software engineering from Cognition V0: Generative UI agent from Vercel GPT Researcher, a research agent  Language Model Cascades: 2022 paper by Google Brain and now OpenAI researcher David Dohan that was influential for Harrison in developing LangChain Transcript: https://www.sequoiacap.com/podcast/training-data-harrison-chase/ 00:00 Introduction 01:21 What are agents?  05:00 What is LangChain's role in the agent ecosystem? 11:13 What is a cognitive architecture?  13:20 Is bespoke and hard coded the way the world is going, or a stop gap? 18:48 Focus on what makes your beer taste better 20:37 So what?  22:20 Where are agents getting traction? 25:35 Reflection, chain of thought, other techniques? 30:42 UX can influence the effectiveness of the architecture 35:30 What's out of scope? 38:04 Fine tuning vs prompting? 42:17 Existing observability tools for LLMs vs needing a new architecture/approach 45:38 Lightning round

Nick Frosst, co-founder of Cohere, on the future of LLMs, and AGI. Learn how Cohere is solving real problems for business with their new AI models. This is the first podcast from our new Cohere partnership! Nick talks about his journey at Google Brain, working with AI legends like Geoff Hinton, and the amazing things his company, Cohere, is doing. From creating the must useful language models for businesses to making tools for developers, Nick shares a lot of interesting insights. He even talks about his band, Good Kid! Nick said that RAG is one of the best features of Cohere's new Command R* models. We are about to release a deep-dive on RAG with Patrick Lewis from Cohere, keep an eye out for that - he explains why their models are specifically optimised for RAG use cases. Learn more about Cohere Command R* models here: https://cohere.com/commandhttps://github.com/cohere-ai/cohere-toolkit Nick's band Good Kid: https://goodkidofficial.com/ Nick on Twitter: https://x.com/nickfrosst Disclaimer: We are in a partnership with Cohere to release content for them. We were not told what to say in the interview, and didn't edit anything out from the interview. We are currently planning to release 2 shows per month under the partnership about their AI platform, research and strategy.

Welcome to the What's Next! Podcast with Tiffani Bova.    I have a special treat for this show. We have not one, but actually two guests.    The first is Martin Gonzalez. He is the creator of Google's Effective Founders Project, a global research program that decodes the factors that enable startup founders to succeed. He also works closely with Google's engineering and research leaders on org design, leadership, and culture challenges.   Joining him is Josh Yellin, who co-founded Google's first Startup Accelerator and spearheaded its growth, reaching founders in 70 countries. Along with Martin, he co-founded Google's Effective Founders Project, and he recently spent four years as Chief of Staff at Google Brain and is presently an organizational leader at Google DeepMind.    Martin and Josh are the authors of a new book called The Bonfire Moment.    THIS EPISODE IS PERFECT FOR… founders of startups who want to avoid common pitfalls.    TODAY'S MAIN MESSAGE… research shows that 65% of startups will fail due to people issues - not the product development, resourcing, or any of the other challenging parts of being a founder. In this episode, Martin and Josh share how they face the people issues head-on and have trained thousands of founders through Bonfire Moment workshops.    Key takeaways:  Startups thrive when people issues are addressed alongside product and market challenges Structured reflection helps startups solve hidden problems Addressing people issues early on fosters long-term growth and stability Finding a co-founder should be a slow and deliberate decision Self-awareness is a critical character trait for founders to develop   WHAT I LOVE MOST… Josh and Martin's recommendation to not using the “maverick mindset” on the organizational side of things. Lean into innovative ideas for your product of service but rely on best practices when approaching leadership.    Running Time: 30:34   Subscribe on iTunes   Find Tiffani Online: Facebook Twitter LinkedIn   Learn More About Martin and Josh:  The Bonfire Moment Website   Book: The Bonfire Moment  

Episode 126I spoke with Vivek Natarajan about:* Improving access to medical knowledge with AI* How an LLM for medicine should behave* Aspects of training Med-PaLM and AMIE* How to facilitate appropriate amounts of trust in users of medical AI systemsVivek Natarajan is a Research Scientist at Google Health AI advancing biomedical AI to help scale world class healthcare to everyone. Vivek is particularly interested in building large language models and multimodal foundation models for biomedical applications and leads the Google Brain moonshot behind Med-PaLM, Google's flagship medical large language model. Med-PaLM has been featured in The Scientific American, The Economist, STAT News, CNBC, Forbes, New Scientist among others.I spend a lot of time on this podcast—if you like my work, you can support me on Patreon :)Reach me at editor@thegradient.pub for feedback, ideas, guest suggestions. Subscribe to The Gradient Podcast:  Apple Podcasts  | Spotify | Pocket Casts | RSSFollow The Gradient on TwitterOutline:* (00:00) Intro* (00:35) The concept of an “AI doctor”* (06:54) Accessibility to medical expertise* (10:31) Enabling doctors to do better/different work* (14:35) Med-PaLM* (15:30) Instruction tuning, desirable traits in LLMs for medicine* (23:41) Axes for evaluation of medical QA systems* (30:03) Medical LLMs and scientific consensus* (35:32) Demographic data and patient interventions* (40:14) Data contamination in Med-PaLM* (42:45) Grounded claims about capabilities* (45:48) Building trust* (50:54) Genetic Discovery enabled by a LLM* (51:33) Novel hypotheses in genetic discovery* (57:10) Levels of abstraction for hypotheses* (1:01:10) Directions for continued progress* (1:03:05) Conversational Diagnostic AI* (1:03:30) Objective Structures Clinical Examination as an evaluative framework* (1:09:08) Relative importance of different types of data* (1:13:52) Self-play — conversational dispositions and handling patients* (1:16:41) Chain of reasoning and information retention* (1:20:00) Performance in different areas of medical expertise* (1:22:35) Towards accurate differential diagnosis* (1:31:40) Feedback mechanisms and expertise, disagreement among clinicians* (1:35:26) Studying trust, user interfaces* (1:38:08) Self-trust in using medical AI models* (1:41:39) UI for medical AI systems* (1:43:50) Model reasoning in complex scenarios* (1:46:33) Prompting* (1:48:41) Future outlooks* (1:54:53) OutroLinks:* Vivek's Twitter and homepage* Papers* Towards Expert-Level Medical Question Answering with LLMs (2023)* LLMs encode clinical knowledge (2023)* Towards Generalist Biomedical AI (2024)* AMIE* Genetic Discovery enabled by a LLM (2023) Get full access to The Gradient at thegradientpub.substack.com/subscribe

Welcome to an interview with Martin Gonzalez and Josh Yellin. Martin is the creator of Google's Effective Founders Project and a frequent lecturer at Stanford, Wharton and INSEAD. Josh co-founded Google's first Startup Accelerator and is presently an organizational leader at Google DeepMind.   Josh and Martin are the authors of The Bonfire Moment: Bring Your Team Together to Solve the Hardest Problems Startups Face. In this book Martin Gonzalez and Josh Yellin outline the common traps startup teams fall into, and share their powerful one-day workshop that helps teams escape those traps. The unique process of The Bonfire Moment brings colleagues together for a full day of facing hard truths, noticing hidden dynamics, and gearing up for the intense challenges of startup life. When the constant hustle feels overwhelming, a team's Bonfire Moment pulls them out of the day-to-day intensity to reflect and reboot.   Martin Gonzalez is the creator of Google's Effective Founders Project, a global research program that decodes the factors that enable startup founders to succeed. He also works closely with Google's engineering and research leaders on org design, leadership and culture challenges. Martin is a frequent lecturer at Stanford, Wharton and INSEAD, and has advised leaders across the Americas, Europe, Africa and Asia. He has been recognized as a Fellow by the Aspen Institute's First Movers Program and as a Thinkers50 Radar awardee, both for his contributions to shaping the future of management and leadership. He studied organizational psychology and behavioral science at Columbia University and the London School of Economics.   Josh Yellin co-founded Google's first Startup Accelerator and spearheaded its growth, reaching founders in 70 countries. Along with Martin, he co-founded Google's Effective Founders Project. Josh recently spent four years as the Chief of Staff at Google Brain, and is presently an organizational leader at Google DeepMind. Josh studied biology at the University of Illinois and business at the Wharton School.   Get The Bonfire Moment here: https://rb.gy/lz7vud   Here are some free gifts for you: Overall Approach Used in Well-Managed Strategy Studies free download: www.firmsconsulting.com/OverallApproach   McKinsey & BCG winning resume free download: www.firmsconsulting.com/resumepdf   Enjoying this episode? Get access to sample advanced training episodes here: www.firmsconsulting.com/promo

In this episode of the Crazy Wisdom Podcast, Stewart Alsop talks with John Ballentine, the founder and CEO of Alchemy.ai. With over seven years of experience in machine learning and large language models (LLMs), John shares insights on synthetic data, the evolution of AI from Google's BERT model to OpenAI's GPT-3, and the future of multimodal algorithms. They discuss the significance of synthetic data in reducing costs and energy for training models, the challenges of creating models that understand natural language, and the exciting potential of AI in various fields, including cybersecurity and creative arts. For more information on John and his work, visit Alchemy.ai.  Check out this GPT we trained on the conversation! Timestamps 00:00 - Stewart Alsop introduces Jon Ballentine, founder and CEO of Alchemy.ai, discussing Jon's background in machine learning and LLMs. 05:00 - Jon talks about the beginnings of his work with the BERT model and the development of transformer architecture. 10:00 - Discussion on the capabilities of early AI models and how they evolved, particularly focusing on the Google Brain project and OpenAI's GPT-3. 15:00 - Exploration of synthetic data, its importance, and how it helps in reducing the cost and energy required for training AI models. 20:00 - Jon discusses the impact of synthetic data on the control and quality of AI model outputs, including challenges and limitations. 25:00 - Conversation about the future of AI, multimodal models, and the significance of video data in training models. 30:00 - The potential of AI in creative fields, such as art, and the concept of artists creating personalized AI models. 35:00 - Challenges in the AI field, including cybersecurity risks and the need for better interpretability of models. 40:00 - The role of synthetic data in enhancing AI training and the discussion on novel attention mechanisms and their applications. 45:00 - Stewart and Jon discuss the relationship between AI and mental health, focusing on therapy and support tools for healthcare providers. 50:00 - The importance of clean data and the challenges of reducing bias and toxicity in AI models, as well as potential future developments in AI ethics and governance. 55:00 - Jon shares more about Alchemy.ai and its mission, along with final thoughts on the future of AI and its societal impacts. Key Insights Evolution of AI Models: Jon Ballentine discusses the evolution of AI models, starting from Google's BERT model to OpenAI's GPT-3. He explains how these models expanded on autocomplete algorithms to predict the next token, with GPT-3 scaling up significantly in parameters and compute. This progression highlights the rapid advancements in natural language processing and the increasing capabilities of AI. Importance of Synthetic Data: Synthetic data is a major focus, with Jon emphasizing its potential to reduce the costs and energy associated with training AI models. He explains that synthetic data allows for better control over model outputs, ensuring that models are trained on diverse and comprehensive datasets without the need for massive amounts of real-world data, which can be expensive and time-consuming to collect. Multimodal Models and Video Data: Jon touches on the importance of multimodal models, which integrate multiple types of data such as text, images, and video. He highlights the potential of video data in training AI models, noting that companies like Google and OpenAI are leveraging vast amounts of video data to improve model performance and capabilities. This approach provides models with a richer understanding of the world from different angles and movements. AI in Creative Fields: The conversation delves into the intersection of AI and creativity. Jon envisions a future where artists create personalized AI models that produce content in their unique style, making art more accessible and personalized. This radical idea suggests that AI could become a new medium for artistic expression, blending technology and creativity in unprecedented ways. Challenges in AI Interpretability: Jon highlights the challenges of understanding and interpreting large AI models. He mentions that despite being able to see the parameters, the internal workings of these models remain largely a black box. This lack of interpretability poses significant challenges, especially in ensuring the safety and reliability of AI systems as they become more integrated into various aspects of life. Cybersecurity Risks and AI: The episode covers the potential cybersecurity risks posed by advanced AI models. Jon discusses the dangers of rogue AI systems that could hack and exfiltrate data, creating new types of cyber threats. This underscores the need for robust cybersecurity measures and the development of defensive AI models to counteract these risks. Future of AI and Mental Health: Stewart and Jon explore the potential of AI in the field of mental health, particularly in supporting healthcare providers. While Jon is skeptical about AI replacing human therapists, he sees value in AI tools that enhance the ability of therapists and doctors to access relevant information and provide better care. This highlights a future where AI augments human capabilities, improving the efficiency and effectiveness of mental health care.

Welcome to Strategy Skills episode 448 with Martin Gonzalez and Josh Yellin. Martin is the creator of Google's Effective Founders Project and a frequent lecturer at Stanford, Wharton and INSEAD. Josh co-founded Google's first Startup Accelerator and is presently an organizational leader at Google DeepMind.   Josh and Martin are the authors of The Bonfire Moment: Bring Your Team Together to Solve the Hardest Problems Startups Face. In this book Martin Gonzalez and Josh Yellin outline the common traps startup teams fall into, and share their powerful one-day workshop that helps teams escape those traps. The unique process of The Bonfire Moment brings colleagues together for a full day of facing hard truths, noticing hidden dynamics, and gearing up for the intense challenges of startup life. When the constant hustle feels overwhelming, a team's Bonfire Moment pulls them out of the day-to-day intensity to reflect and reboot.   Martin Gonzalez is the creator of Google's Effective Founders Project, a global research program that decodes the factors that enable startup founders to succeed. He also works closely with Google's engineering and research leaders on org design, leadership and culture challenges. Martin is a frequent lecturer at Stanford, Wharton and INSEAD, and has advised leaders across the Americas, Europe, Africa and Asia. He has been recognized as a Fellow by the Aspen Institute's First Movers Program and as a Thinkers50 Radar awardee, both for his contributions to shaping the future of management and leadership. He studied organizational psychology and behavioral science at Columbia University and the London School of Economics.   Josh Yellin co-founded Google's first Startup Accelerator and spearheaded its growth, reaching founders in 70 countries. Along with Martin, he co-founded Google's Effective Founders Project. Josh recently spent four years as the Chief of Staff at Google Brain, and is presently an organizational leader at Google DeepMind. Josh studied biology at the University of Illinois and business at the Wharton School.   Get The Bonfire Moment here: https://rb.gy/lz7vud   Here are some free gifts for you: Overall Approach Used in Well-Managed Strategy Studies free download: www.firmsconsulting.com/OverallApproach   McKinsey & BCG winning resume free download: www.firmsconsulting.com/resumepdf   Enjoying this episode? Get access to sample advanced training episodes here: www.firmsconsulting.com/promo

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Should we break up Google DeepMind?, published by Hauke Hillebrandt on April 23, 2024 on The Effective Altruism Forum. Regulators should review the 2014 DeepMind acquisition. When Google bought DeepMind in 2014, no regulator, not the FTC, not the EC's DG COMP, nor the CMA, scrutinized the impact. Why? AI startups have high value but low revenues. And so they avoid regulation (and tax, see below). Buying start-ups with low revenues flies under the thresholds of EU merger regulation[1] or the CMA's 'turnover test' (despite it being a 'relevant enterprise' under the National Security and Investment Act). In 2020, the FTC ordered Big Tech to provide info on M&A from 2010-2019 that it didn't report (UK regulators should urgently do so as well given that their retrospective powers might only be 10 years).[2] Regulators should also review the 2023 Google-DeepMind internal merger. DeepMind and Google Brain are key players in AI. In 2023, they merged into Google DeepMind. This compromises independence, reduces competition for AI talent and resources, and limits alternatives for collaboration partners. Though they are both part of Google, regulators can scrutinize this, regardless of corporate structure. For instance, UK regulators have intervened in M&A of enterprises already under common ownership - especially in Tech (cf UK regulators ordered FB to sell GIPHY). And so, regulators should consider breaking up Google Deepmind as per recent proposals: A new paper 'Unscrambling the eggs: breaking up consummated mergers and dominant firms' by economists at Imperial cites Google DeepMind as a firm that could be unmerged. [3] A new Brookings paper also argues that if other means to ensure fair markets fail, then as a last resort, foundation model firms may need to be broken up on the basis of functions, akin to how we broke up AT&T.[4] Relatedly, some top economists agree that we should designate Google Search as 'platform utilities' and break it apart from any participant on that platform, most agree that we should explore this further to weigh costs and benefits.[5] Indeed, the EU accuses Google of abusing dominance in ad tech and may force it to sell parts of its firm.[6] Kustomer, a firm of a similar size to DeepMind bought by Facebook, recently spun out again and shows this is possible. Finally, DeepMind itself has in the past tried to break away from Google.[7] Since DeepMind's AI improves all Google products, regulators should work cross-departmentally to scrutinize both mergers above on the following grounds: Market dominance: Google dominates the field of AI, surpassing all universities in terms of high-quality publications: Tax avoidance: Despite billions in UK profits yearly, Google is only taxed $60M.[8] DeepMind's is only taxed ~$1M per year on average. [9],[10] We should tax them more fairly. DeepMind's recent revenue jump is due to creative accounting, as it doesn't have many revenue streams, but almost all are based on how much Google arbitrarily pays for internal services. Indeed, Google just waived $1.5B in DeepMind's 'startup debt' [11],[12] despite DeepMind's CEO boasting that they have a unique opportunity as part of Google and its dozens of billion user products by immediately shipping their advances into[13] and saving Google hundreds of millions in energy costs.[14] About 85% of the innovations causing the recent AI boom came from Google DeepMind.[15] DeepMind also holds 560 patents,[16] and this IP is very hard to value and tax. Such a bad precedent might cause either more tax avoidance by OpenAI, Microsoft AI, Anthropic, Palantir, and A16z setting up UK offices, or it will give Google an unfair edge over these smaller firms). Public interest concerns: DeepMind's AI improves YouTube's algorithm and thus DeepMind indirectly polarizes voters.[17] Regulators s...

Dr. Andrew Ning: A prominent computer scientist known for his contributions to artificial intelligence. Co-founder and former head of Google Brain, ex-chief scientist at Baidu, co-founder of Coursera. His educational background includes UC Berkeley, MIT, and Carnegie Mellon.Topic: Ning's talk at Sequoia Capital focused on agents and their potential in AI, advocating the powerful capabilities of agents when powered by models like GPT-3.5, asserting they can perform at the level of GPT-4.Video: https://youtu.be/ZYf9V2fSFwU?si=7gSAHfJSiGkuvNEK

Today we bring you a very special episode of Crosscurrents. First, we hear how the coastal town of San Pippo sees climate change as a glass half full. Then, we bring you practical advice on combating airline anxiety. And, we have conversation with the very first user of the new Google Brain.

Our next SF event is AI UX 2024 - let's see the new frontier for UX since last year! Last call: we are recording a preview of the AI Engineer World's Fair with swyx and Ben Dunphy, send any questions about Speaker CFPs and Sponsor Guides you have!Alessio is now hiring engineers for a new startup he is incubating at Decibel: Ideal candidate is an “ex-technical co-founder type”. Reach out to him for more!David Luan has been at the center of the modern AI revolution: he was the ~30th hire at OpenAI, he led Google's LLM efforts and co-led Google Brain, and then started Adept in 2022, one of the leading companies in the AI agents space. In today's episode, we asked David for some war stories from his time in early OpenAI (including working with Alec Radford ahead of the GPT-2 demo with Sam Altman, that resulted in Microsoft's initial $1b investment), and how Adept is building agents that can “do anything a human does on a computer" — his definition of useful AGI.Why Google *couldn't* make GPT-3While we wanted to discuss Adept, we couldn't talk to a former VP Eng of OpenAI and former LLM tech lead at Google Brain and not ask about the elephant in the room. It's often asked how Google had such a huge lead in 2017 with Vaswani et al creating the Transformer and Noam Shazeer predicting trillion-parameter models and yet it was David's team at OpenAI who ended up making GPT 1/2/3. David has some interesting answers:“So I think the real story of GPT starts at Google, of course, right? Because that's where Transformers sort of came about. However, the number one shocking thing to me was that, and this is like a consequence of the way that Google is organized…what they (should) have done would be say, hey, Noam Shazeer, you're a brilliant guy. You know how to scale these things up. Here's half of all of our TPUs. And then I think they would have destroyed us. He clearly wanted it too…You know, every day we were scaling up GPT-3, I would wake up and just be stressed. And I was stressed because, you know, you just look at the facts, right? Google has all this compute. Google has all the people who invented all of these underlying technologies. There's a guy named Noam who's really smart, who's already gone and done this talk about how he wants a trillion parameter model. And I'm just like, we're probably just doing duplicative research to what he's doing. He's got this decoder only transformer that's probably going to get there before we do. And it turned out the whole time that they just couldn't get critical mass. So during my year where I led the Google LM effort and I was one of the brain leads, you know, it became really clear why. At the time, there was a thing called the Brain Credit Marketplace. Everyone's assigned a credit. So if you have a credit, you get to buy end chips according to supply and demand. So if you want to go do a giant job, you had to convince like 19 or 20 of your colleagues not to do work. And if that's how it works, it's really hard to get that bottom up critical mass to go scale these things. And the team at Google were fighting valiantly, but we were able to beat them simply because we took big swings and we focused.”Cloning HGI for AGIHuman intelligence got to where it is today through evolution. Some argue that to get to AGI, we will approximate all the “FLOPs” that went into that process, an approach most famously mapped out by Ajeya Cotra's Biological Anchors report:The early days of OpenAI were very reinforcement learning-driven with the Dota project, but that's a very inefficient way for these models to re-learn everything. (Kanjun from Imbue shared similar ideas in her episode).David argues that there's a shortcut. We can bootstrap from existing intelligence.“Years ago, I had a debate with a Berkeley professor as to what will it actually take to build AGI. And his view is basically that you have to reproduce all the flops that went into evolution in order to be able to get there… I think we are ignoring the fact that you have a giant shortcut, which is you can behaviorally clone everything humans already know. And that's what we solved with LLMs!”LLMs today basically model intelligence using all (good!) written knowledge (see our Datasets 101 episode), and have now expanded to non-verbal knowledge (see our HuggingFace episode on multimodality). The SOTA self-supervised pre-training process is surprisingly data-efficient in taking large amounts of unstructured data, and approximating reasoning without overfitting.But how do you cross the gap from the LLMs of today to building the AGI we all want? This is why David & friends left to start Adept.“We believe the clearest framing of general intelligence is a system that can do anything a human can do in front of a computer. A foundation model for actions, trained to use every software tool, API, and webapp that exists, is a practical path to this ambitious goal” — ACT-1 BlogpostCritical Path: Abstraction with ReliabilityThe AGI dream is fully autonomous agents, but there are levels to autonomy that we are comfortable giving our agents, based on how reliable they are. In David's word choice, we always want higher levels of “abstractions” (aka autonomy), but our need for “reliability” is the practical limit on how high of an abstraction we can use.“The critical path for Adept is we want to build agents that can do a higher and higher level abstraction things over time, all while keeping an insanely high reliability standard. Because that's what turns us from research into something that customers want. And if you build agents with really high reliability standard, but are continuing pushing a level of abstraction, you then learn from your users how to get that next level of abstraction faster. So that's how you actually build the data flow. That's the critical path for the company. Everything we do is in service of that.”We saw how Adept thinks about different levels of abstraction at the 2023 Summit:The highest abstraction is the “AI Employee”, but we'll get there with “AI enabled employees”. Alessio recently gave a talk about the future of work with “services as software” at this week's Nvidia GTC (slides).No APIsUnlike a lot of large research labs, Adept's framing of AGI as "being able to use your computer like a human" carries with it a useful environmental constraint:“Having a human robot lets you do things that humans do without changing everything along the way. It's the same thing for software, right? If you go itemize out the number of things you want to do on your computer for which every step has an API, those numbers of workflows add up pretty close to zero. And so then many points along the way, you need the ability to actually control your computer like a human. It also lets you learn from human usage of computers as a source of training data that you don't get if you have to somehow figure out how every particular step needs to be some particular custom private API thing. And so I think this is actually the most practical path (to economic value).”This realization and conviction means that multimodal modals are the way to go. Instead of using function calling to call APIs to build agents, which is what OpenAI and most of the open LLM industry have done to date, Adept wants to “drive by vision”, (aka see the screen as a human sees it) and pinpoint where to click and type as a human does. No APIs needed, because most software don't expose APIs.Extra context for readers: You can see the DeepMind SIMA model in the same light: One system that learned to play a diverse set of games (instead of one dedicated model per game) using only pixel inputs and keyboard-and-mouse action outputs!The OpenInterpreter team is working on a “Computer API” that also does the same.To do this, Adept had to double down on a special kind of multimodality for knowledge work:“A giant thing that was really necessary is really fast multimodal models that are really good at understanding knowledge work and really good at understanding screens. And that is needs to kind of be the base for some of these agents……I think one big hangover primarily academic focus for multimodal models is most multimodal models are primarily trained on like natural images, cat and dog photos, stuff that's come out of the camera… (but) where are they going to be the most useful? They're going to be most useful in knowledge work tasks. That's where the majority of economic value is going to be. It's not in cat and dogs. And so if that's what it is, what do you need to train? I need to train on like charts, graphs, tables, invoices, PDFs, receipts, unstructured data, UIs. That's just a totally different pre-training corpus. And so Adept spent a lot of time building that.”With this context, you can now understand the full path of Adept's public releases:* ACT-1 (Sept 2022): a large Transformers model optimized for browser interactions. It has a custom rendering of the browser viewport that allows it to better understand it and take actions.* Persimmon-8B (Sept 2023): a permissive open LLM (weights and code here)* Fuyu-8B (Oct 2023): a small version of the multimodal model that powers Adept. Vanilla decoder-only transformer with no specialized image encoder, which allows it to handle input images of varying resolutions without downsampling.* Adept Experiments (Nov 2023): A public tool to build automations in the browser. This is powered by Adept's core technology but it's just a piece of their enterprise platform. They use it as a way to try various design ideas.* Fuyu Heavy (Jan 2024) - a new multimodal model designed specifically for digital agents and the world's third-most-capable multimodal model (beating Gemini Pro on MMMU, AI2D, and ChartQA), “behind only GPT4-V and Gemini Ultra, which are 10-20 times bigger”The Fuyu-8B post in particular exhibits a great number of examples on knowledge work multimodality:Why Adept is NOT a Research LabWith OpenAI now worth >$90b and Anthropic >$18b, it is tempting to conclude that the AI startup metagame is to build a large research lab, and attract the brightest minds and highest capital to build AGI. Our past guests (see the Humanloop episode) and (from Imbue) combined to ask the most challenging questions of the pod - with David/Adept's deep research pedigree from Deepmind and OpenAI, why is Adept not building more general foundation models (like Persimmon) and playing the academic benchmarks game? Why is Adept so focused on commercial agents instead?“I feel super good that we're doing foundation models in service of agents and all of the reward within Adept is flowing from “Can we make a better agent”…… I think pure play foundation model companies are just going to be pinched by how good the next couple of (Meta Llama models) are going to be… And then seeing the really big players put ridiculous amounts of compute behind just training these base foundation models, I think is going to commoditize a lot of the regular LLMs and soon regular multimodal models. So I feel really good that we're just focused on agents.”and the commercial grounding is his answer to Kanjun too (whom we also asked the inverse question to compare with Adept):“… the second reason I work at Adept is if you believe that actually having customers and a reward signal from customers lets you build AGI faster, which we really believe, then you should come here. And I think the examples for why that's true is for example, our evaluations are not academic evals. They're not simulator evals. They're like, okay, we have a customer that really needs us to do these particular things. We can do some of them. These are the ones they want us to, we can't do them at all. We've turned those into evals.. I think that's a degree of practicality that really helps.”And his customers seem pretty happy, because David didn't need to come on to do a sales pitch:David: “One of the things we haven't shared before is we're completely sold out for Q1.”Swyx: “Sold out of what?”David: “Sold out of bandwidth to onboard more customers.”Well, that's a great problem to have.Show Notes* David Luan* Dextro at Data Driven NYC (2015)* Adept* ACT-1* Persimmon-8B* Adept Experiments* Fuyu-8B* $350M Series B announcement* Amelia Wattenberger talk at AI Engineer Summit* FigureChapters* [00:00:00] Introductions* [00:01:14] Being employee #30 at OpenAI and its early days* [00:13:38] What is Adept and how do you define AGI?* [00:21:00] Adept's critical path and research directions* [00:26:23] How AI agents should interact with software and impact product development* [00:30:37] Analogies between AI agents and self-driving car development* [00:32:42] Balancing reliability, cost, speed and generality in AI agents* [00:37:30] Potential of foundation models for robotics* [00:39:22] Core research questions and reasons to work at AdeptTranscriptsAlessio [00:00:00]: Hey everyone, welcome to the Latent Space Podcast. This is Alessio, partner and CTO in Residence at Decibel Partners, and I'm joined by my co-host Swyx, founder of Smol.ai.Swyx [00:00:15]: Hey, and today we have David Luan, CEO, co-founder of Adept in the studio. Welcome.David [00:00:20]: Yeah, thanks for having me.Swyx [00:00:21]: Been a while in the works. I've met you socially at one of those VC events and you said that you were interested in coming on and glad we finally were able to make this happen.David: Yeah, happy to be part of it.Swyx: So we like to introduce the speaker and then also just like have you talk a little bit about like what's not on your LinkedIn, what people should just generally know about you. You started a company in college, which was the first sort of real time video detection classification API that was Dextro, and that was your route to getting acquired into Axon where you're a director of AI. Then you were the 30th hire at OpenAI?David [00:00:53]: Yeah, 30, 35, something around there. Something like that.Swyx [00:00:56]: So you were VP of Eng for two and a half years to two years, briefly served as tech lead of large models at Google, and then in 2022 started Adept. So that's the sort of brief CV. Is there anything else you like want to fill in the blanks or like people should know more about?David [00:01:14]: I guess a broader story was I joined OpenAI fairly early and I did that for about two and a half to three years leading engineering there. It's really funny, I think second or third day of my time at OpenAI, Greg and Ilya pulled me in a room and we're like, you know, you should take over our directs and we'll go mostly do IC work. So that was fun, just coalescing a bunch of teams out of a couple of early initiatives that had already happened. The company, the Dota effort was going pretty hard and then more broadly trying to put bigger picture direction around what we were doing with basic research. So I spent a lot of time doing that. And then I led Google's LLM efforts, but also co-led Google Brain was one of the brain leads more broadly. You know, there's been a couple of different eras of AI research, right? If we count everything before 2012 as prehistory, which people hate it when I say that, kind of had this like you and your three best friends write a research paper that changes the world period from like 2012 to 2017. And I think the game changed in 2017 and like most labs didn't realize it, but we at OpenAI really did. I think in large part helped by like Ilya's constant beating of the drum that the world would be covered in data centers. And I think-Swyx [00:02:15]: It's causally neat.David [00:02:16]: Yeah. Well, like I think we had conviction in that, but it wasn't until we started seeing results that it became clear that that was where we had to go. But also part of it as well was for OpenAI, like when I first joined, I think one of the jobs that I had to do was how do I tell a differentiated vision for who we were technically compared to, you know, hey, we're just smaller Google Brain, or like you work at OpenAI if you live in SF and don't want to commute to Mountain View or don't want to live in London, right? That's like not enough to like hang your technical identity as a company. And so what we really did was, and I spent a lot of time pushing this, is just how do we get ourselves focused on a certain class of like giant swings and bets, right? Like how do you flip the script from you just do bottom-up research to more about how do you like leave some room for that, but really make it about like, what are the big scientific outcomes that you want to show? And then you just solve them at all costs, whether or not you care about novelty and all that stuff. And that became the dominant model for a couple of years, right? And then what's changed now is I think the number one driver of AI products over the next couple of years is going to be the deep co-design and co-evolution of product and users for feedback and actual technology. And I think labs, every tool to go do that are going to do really well. And that's a big part of why I started Adept.Alessio [00:03:20]: You mentioned Dota, any memories thinking from like the switch from RL to Transformers at the time and kind of how the industry was evolving more in the LLM side and leaving behind some of the more agent simulation work?David [00:03:33]: Like zooming way out, I think agents are just absolutely the correct long-term direction, right? You just go to find what AGI is, right? You're like, Hey, like, well, first off, actually, I don't love AGI definitions that involve human replacement because I don't think that's actually how it's going to happen. Even this definition of like, Hey, AGI is something that outperforms humans at economically valuable tasks is kind of implicit view of the world about what's going to be the role of people. I think what I'm more interested in is like a definition of AGI that's oriented around like a model that can do anything a human can do on a computer. If you go think about that, which is like super tractable, then agent is just a natural consequence of that definition. And so what did all the work we did on our own stuff like that get us was it got us a really clear formulation. Like you have a goal and you want to maximize the goal, you want to maximize reward, right? And the natural LLM formulation doesn't come with that out of the box, right? I think that we as a field got a lot right by thinking about, Hey, how do we solve problems of that caliber? And then the thing we forgot is the Novo RL is like a pretty terrible way to get there quickly. Why are we rediscovering all the knowledge about the world? Years ago, I had a debate with a Berkeley professor as to what will it actually take to build AGI. And his view is basically that you have to reproduce all the flops that went into evolution in order to be able to get there. Right.Swyx [00:04:44]: The biological basis theory. Right.David [00:04:46]: So I think we are ignoring the fact that you have a giant shortcut, which is you can behavioral clone everything humans already know. And that's what we solved with LLMs. We've solved behavioral cloning, everything that humans already know. Right. So like today, maybe LLMs is like behavioral cloning every word that gets written on the internet in the future, the multimodal models are becoming more of a thing where behavioral cloning the visual world. But really, what we're just going to have is like a universal byte model, right? Where tokens of data that have high signal come in, and then all of those patterns are like learned by the model. And then you can regurgitate any combination now. Right. So text into voice out, like image into other image out or video out or whatever, like these like mappings, right? Like all just going to be learned by this universal behavioral cloner. And so I'm glad we figured that out. And I think now we're back to the era of how do we combine this with all of the lessons we learned during the RL period. That's what's going to drive progress.Swyx [00:05:35]: I'm still going to pressure you for a few more early opening stories before we turn to the ADET stuff. On your personal site, which I love, because it's really nice, like personal, you know, story context around like your history. I need to update it. It's so old. Yeah, it's so out of date. But you mentioned GPT-2. Did you overlap with GPT-1? I think you did, right?David [00:05:53]: I actually don't quite remember. I think I was joining right around- Right around then?Swyx [00:05:57]: I was right around that, yeah. Yeah. So what I remember was Alec, you know, just kind of came in and was like very obsessed with Transformers and applying them to like Reddit sentiment analysis. Yeah, sentiment, that's right. Take us through-David [00:06:09]: Sentiment neuron, all this stuff.Swyx [00:06:10]: The history of GPT as far as you know, you know, according to you. Ah, okay.David [00:06:14]: History of GPT, according to me, that's a pretty good question. So I think the real story of GPT starts at Google, of course, right? Because that's where Transformers sort of came about. However, the number one shocking thing to me was that, and this is like a consequence of the way that Google is organized, where like, again, you and your three best friends write papers, right? Okay. So zooming way out, right? I think about my job when I was a full-time research leader as a little bit of a portfolio allocator, right? So I've got really, really smart people. My job is to convince people to coalesce around a small number of really good ideas and then run them over the finish line. My job is not actually to promote a million ideas and never have critical mass. And then as the ideas start coming together and some of them start working well, my job is to nudge resources towards the things that are really working and then start disbanding some of the things that are not working, right? That muscle did not exist during my time at Google. And I think had they had it, what they would have done would be say, hey, Noam Shazir, you're a brilliant guy. You know how to scale these things up. Here's half of all of our TPUs. And then I think they would have destroyed us. He clearly wanted it too.Swyx [00:07:17]: He's talking about trillion parameter models in 2017.David [00:07:20]: Yeah. So that's the core of the GPT story, right? Which is that, and I'm jumping around historically, right? But after GPT-2, we were all really excited about GPT-2. I can tell you more stories about that. It was the last paper that I even got to really touch before everything became more about building a research org. You know, every day we were scaling up GPT-3, I would wake up and just be stressed. And I was stressed because, you know, you just look at the facts, right? Google has all this compute. Google has all the people who invented all of these underlying technologies. There's a guy named Noam who's really smart, who's already gone and done this talk about how he wants a trillion parameter model. And I'm just like, we're probably just doing duplicative research to what he's doing, right? He's got this decoder only transformer that's probably going to get there before we do. And I was like, but like, please just like let this model finish, right? And it turned out the whole time that they just couldn't get critical mass. So during my year where I led the Google LM effort and I was one of the brain leads, you know, it became really clear why, right? At the time, there was a thing called the brain credit marketplace. And did you guys know the brain credit marketplace? No, I never heard of this. Oh, so it's actually, it's a, you can ask any Googler.Swyx [00:08:23]: It's like just like a thing that, that, I mean, look like, yeah, limited resources, you got to have some kind of marketplace, right? You know, sometimes it's explicit, sometimes it isn't, you know, just political favors.David [00:08:34]: You could. And so then basically everyone's assigned a credit, right? So if you have a credit, you get to buy end chips according to supply and demand. So if you want to go do a giant job, you had to convince like 19 or 20 of your colleagues not to do work. And if that's how it works, it's really hard to get that bottom up critical mass to go scale these things. And the team at Google were fighting valiantly, but we were able to beat them simply because we took big swings and we focused. And I think, again, that's like part of the narrative of like this phase one of AI, right? Of like this modern AI era to phase two. And I think in the same way, I think phase three company is going to out execute phase two companies because of the same asymmetry of success.Swyx [00:09:12]: Yeah. I think it's underrated how much NVIDIA works with you in the early days as well. I think maybe, I think it was Jensen. I'm not sure who circulated a recent photo of him delivering the first DGX to you guys.David [00:09:24]: I think Jensen has been a complete legend and a mastermind throughout. I have so much respect for NVIDIA. It is unreal.Swyx [00:09:34]: But like with OpenAI, like kind of give their requirements, like co-design it or just work of whatever NVIDIA gave them.David [00:09:40]: So we work really closely with them. There's, I'm not sure I can share all the stories, but examples of ones that I've found particularly interesting. So Scott Gray is amazing. I really like working with him. He was on one of my teams, the supercomputing team, which Chris Berner runs and Chris Berner still does a lot of stuff in that. As a result, like we had very close ties to NVIDIA. Actually, one of my co-founders at Adept, Eric Elson, was also one of the early GPGPU people. So he and Scott and Brian Catanzaro at NVIDIA and Jonah and Ian at NVIDIA, I think all were very close. And we're all sort of part of this group of how do we push these chips to the absolute limit? And I think that kind of collaboration helped quite a bit. I think one interesting set of stuff is knowing the A100 generation, that like quad sparsity was going to be a thing. Is that something that we want to go look into, right? And figure out if that's something that we could actually use for model training. Really what it boils down to is that, and I think more and more people realize this, six years ago, people, even three years ago, people refused to accept it. This era of AI is really a story of compute. It's really the story of how do you more efficiently map actual usable model flops to compute,Swyx [00:10:38]: Is there another GPT 2, 3 story that you love to get out there that you think is underappreciated for the amount of work that people put into it?David [00:10:48]: So two interesting GPT 2 stories. One of them was I spent a good bit of time just sprinting to help Alec get the paper out. And I remember one of the most entertaining moments was we were writing the modeling section. And I'm pretty sure the modeling section was the shortest modeling section of any ML, reasonably legitimate ML paper to that moment. It was like section three model. This is a standard vanilla decoder only transformer with like these particular things, those paragraph long if I remember correctly. And both of us were just looking at the same being like, man, the OGs in the field are going to hate this. They're going to say no novelty. Why did you guys do this work? So now it's funny to look at in hindsight that it was pivotal kind of paper, but I think it was one of the early ones where we just leaned fully into all we care about is solving problems in AI and not about, hey, is there like four different really simple ideas that are cloaked in mathematical language that doesn't actually help move the field forward?Swyx [00:11:42]: Right. And it's like you innovate on maybe like data set and scaling and not so much the architecture.David [00:11:48]: We all know how it works now, right? Which is that there's a collection of really hard won knowledge that you get only by being at the frontiers of scale. And that hard won knowledge, a lot of it's not published. A lot of it is stuff that's actually not even easily reducible to what looks like a typical academic paper. But yet that's the stuff that helps differentiate one scaling program from another. You had a second one? So the second one is, there's like some details here that I probably shouldn't fully share, but hilariously enough for the last meeting we did with Microsoft before Microsoft invested in OpenAI, Sam Altman, myself and our CFO flew up to Seattle to do the final pitch meeting. And I'd been a founder before. So I always had a tremendous amount of anxiety about partner meetings, which this basically this is what it was. I had Kevin Scott and Satya and Amy Hood, and it was my job to give the technical slides about what's the path to AGI, what's our research portfolio, all of this stuff, but it was also my job to give the GPT-2 demo. We had a slightly bigger version of GPT-2 that we had just cut maybe a day or two before this flight up. And as we all know now, model behaviors you find predictable at one checkpoint are not predictable in another checkpoint. And so I'd spent all this time trying to figure out how to keep this thing on rails. I had my canned demos, but I knew I had to go turn it around over to Satya and Kevin and let them type anything in. And that just, that really kept me up all night.Swyx [00:13:06]: Nice. Yeah.Alessio [00:13:08]: I mean, that must have helped you talking about partners meeting. You raised $420 million for Adept. The last round was a $350 million Series B, so I'm sure you do great in partner meetings.Swyx [00:13:18]: Pitchers meetings. Nice.David [00:13:20]: No, that's a high compliment coming from a VC.Alessio [00:13:22]: Yeah, no, I mean, you're doing great already for us. Let's talk about Adept. And we were doing pre-prep and you mentioned that maybe a lot of people don't understand what Adept is. So usually we try and introduce the product and then have the founders fill in the blanks, but maybe let's do the reverse. Like what is Adept? Yeah.David [00:13:38]: So I think Adept is the least understood company in the broader space of foundational models plus agents. So I'll give some color and I'll explain what it is and I'll explain also why it's actually pretty different from what people would have guessed. So the goal for Adept is we basically want to build an AI agent that can do, that can basically help humans do anything a human does on a computer. And so what that really means is we want this thing to be super good at turning natural language like goal specifications right into the correct set of end steps and then also have all the correct sensors and actuators to go get that thing done for you across any software tool that you already use. And so the end vision of this is effectively like I think in a couple of years everyone's going to have access to like an AI teammate that they can delegate arbitrary tasks to and then also be able to, you know, use it as a sounding board and just be way, way, way more productive. Right. And just changes the shape of every job from something where you're mostly doing execution to something where you're mostly actually doing like these core liberal arts skills of what should I be doing and why. Right. And I find this like really exciting and motivating because I think it's actually a pretty different vision for how AGI will play out. I think systems like Adept are the most likely systems to be proto-AGIs. But I think the ways in which we are really counterintuitive to everybody is that we've actually been really quiet because we are not a developer company. We don't sell APIs. We don't sell open source models. We also don't sell bottom up products. We're not a thing that you go and click and download the extension and like we want more users signing up for that thing. We're actually an enterprise company. So what we do is we work with a range of different companies, some like late stage multi-thousand people startups, some fortune 500s, et cetera. And what we do for them is we basically give them an out of the box solution where big complex workflows that their employees do every day could be delegated to the model. And so we look a little different from other companies in that in order to go build this full agent thing, the most important thing you got to get right is reliability. So initially zooming way back when, one of the first things that DEP did was we released this demo called Act One, right? Act One was like pretty cool. It's like kind of become a hello world thing for people to show agent demos by going to Redfin and asking to buy a house somewhere because like we did that in the original Act One demo and like showed that, showed like Google Sheets, all this other stuff. Over the last like year since that has come out, there's been a lot of really cool demos and you go play with them and you realize they work 60% of the time. But since we've always been focused on how do we build an amazing enterprise product, enterprises can't use anything that isn't in the nines of reliability. And so we've actually had to go down a slightly different tech tree than what you might find in the prompt engineering sort of plays in the agent space to get that reliability. And we've decided to prioritize reliability over all else. So like one of our use cases is crazy enough that it actually ends with a physical truck being sent to a place as the result of the agent workflow. And if you're like, if that works like 60% of the time, you're just blowing money and poor truck drivers going places.Alessio [00:16:30]: Interesting. One of the, our investment teams has this idea of services as software. I'm actually giving a talk at NVIDIA GTC about this, but basically software as a service, you're wrapping user productivity in software with agents and services as software is replacing things that, you know, you would ask somebody to do and the software just does it for you. When you think about these use cases, do the users still go in and look at the agent kind of like doing the things and can intervene or like are they totally removed from them? Like the truck thing is like, does the truck just show up or are there people in the middle checking in?David [00:17:04]: I think there's two current flaws in the framing for services as software, or I think what you just said. I think that one of them is like in our experience, as we've been rolling out Adept, the people who actually do the jobs are the most excited about it because they don't go from, I do this job to, I don't do this job. They go from, I do this job for everything, including the shitty rote stuff to I'm a supervisor. And I literally like, it's pretty magical when you watch the thing being used because now it parallelizes a bunch of the things that you had to do sequentially by hand as a human. And you can just click into any one of them and be like, Hey, I want to watch the trajectory that the agent went through to go solve this. And the nice thing about agent execution as opposed to like LLM generations is that a good chunk of the time when the agent fails to execute, it doesn't give you the wrong result. It just fails to execute. And the whole trajectory is just broken and dead and the agent knows it, right? So then those are the ones that the human then goes and solves. And so then they become a troubleshooter. They work on the more challenging stuff. They get way, way more stuff done and they're really excited about it. I think the second piece of it that we've found is our strategy as a company is to always be an augmentation company. And I think one out of principle, that's something we really care about. But two, actually, if you're framing yourself as an augmentation company, you're always going to live in a world where you're solving tasks that are a little too hard for what the model can do today and still needs a human to provide oversight, provide clarifications, provide human feedback. And that's how you build a data flywheel. That's how you actually learn from the smartest humans how to solve things models can't do today. And so I actually think that being an augmentation company forces you to go develop your core AI capabilities faster than someone who's saying, ah, okay, my job is to deliver you a lights off solution for X.Alessio [00:18:42]: Yeah. It's interesting because we've seen two parts of the market. One is we have one company that does agents for SOC analysts. People just don't have them, you know, and just they cannot attract the talent to do it. And similarly, in a software development, you have Copilot, which is the augmentation product, and then you have sweep.dev and you have these products, which they just do the whole thing. I'm really curious to see how that evolves. I agree that today the reliability is so important in the enterprise that they just don't use most of them. Yeah. Yeah. No, that's cool. But it's great to hear the story because I think from the outside, people are like, oh, a dev, they do Act One, they do Persimon, they do Fuyu, they do all this stuff. Yeah, it's just the public stuff.Swyx [00:19:20]: It's just public stuff.David [00:19:21]: So one of the things we haven't shared before is we're completely sold out for Q1. And so I think...Swyx [00:19:26]: Sold out of what?David [00:19:27]: Sold out of bandwidth to go on board more customers. And so we're like working really hard to go make that less of a bottleneck, but our expectation is that I think we're going to be significantly more public about the broader product shape and the new types of customers we want to attract later this year. So I think that clarification will happen by default.Swyx [00:19:43]: Why have you become more public? You know, if the whole push has... You're sold out, you're my enterprise, but you're also clearly putting effort towards being more open or releasing more things.David [00:19:53]: I think we just flipped over that way fairly recently. That's a good question. I think it actually boils down to two things. One, I think that, frankly, a big part of it is that the public narrative is really forming around agents as being the most important thing. And I'm really glad that's happening because when we started the company in January 2022, everybody in the field knew about the agents thing from RL, but the general public had no conception of what it was. They were still hanging their narrative hat on the tree of everything's a chatbot. And so I think now one of the things that I really care about is that when people think agent, they actually think the right thing. All sorts of different things are being called agents. Chatbots are being called agents. Things that make a function call are being called agents. To me, an agent is something that you can give a goal and get an end step workflow done correctly in the minimum number of steps. And so that's a big part of why. And I think the other part is because I think it's always good for people to be more aware of Redept as they think about what the next thing they want to do in their careers. The field is quickly pivoting in a world where foundation models are looking more and more commodity. And I think a huge amount of gain is going to happen from how do you use foundation models as the well-learned behavioral cloner to go solve agents. And I think people who want to do agents research should really come to Redept.Swyx [00:21:00]: When you say agents have become more part of the public narrative, are there specific things that you point to? I'll name a few. Bill Gates in his blog post mentioning that agents are the future. I'm the guy who made OSes, and I think agents are the next thing. So Bill Gates, I'll call that out. And then maybe Sam Altman also saying that agents are the future for open AI.David [00:21:17]: I think before that even, I think there was something like the New York Times, Cade Metz wrote a New York Times piece about it. Right now, in a bit to differentiate, I'm seeing AI startups that used to just brand themselves as an AI company, but now brand themselves as an AI agent company. It's just like, it's a term I just feel like people really want.Swyx [00:21:31]: From the VC side, it's a bit mixed. Is it? As in like, I think there are a lot of VCs where like, I would not touch any agent startups because like- Why is that? Well, you tell me.Alessio [00:21:41]: I think a lot of VCs that are maybe less technical don't understand the limitations of the-Swyx [00:21:46]: No, that's not fair.Alessio [00:21:47]: No, no, no, no. I think like- You think so? No, no. I think like the, what is possible today and like what is worth investing in, you know? And I think like, I mean, people look at you and say, well, these guys are building agents. They needed 400 million to do it. So a lot of VCs are maybe like, oh, I would rather invest in something that is tacking on AI to an existing thing, which is like easier to get the market and kind of get some of the flywheel going. But I'm also surprised a lot of funders just don't want to do agents. It's not even the funding. Sometimes we look around and it's like, why is nobody doing agents for X? Wow.David [00:22:17]: That's good to know actually. I never knew that before. My sense from my limited perspective is there's a new agent company popping up every day.Swyx [00:22:24]: So maybe I'm- They are. They are. But like I have advised people to take agents off of their title because it's so diluted.David [00:22:31]: It's now so diluted.Swyx [00:22:32]: Yeah. So then it doesn't stand for anything. Yeah.David [00:22:35]: That's a really good point.Swyx [00:22:36]: So like, you know, you're a portfolio allocator. You have people know about Persimmon, people know about Fuyu and Fuyu Heavy. Can you take us through like how you think about that evolution of that and what people should think about what that means for adepts and sort of research directions? Kind of take us through the stuff you shipped recently and how people should think about the trajectory of what you're doing.David [00:22:56]: The critical path for adepts is we want to build agents that can do a higher and higher level abstraction things over time, all while keeping an insanely high reliability standard. Because that's what turns us from research into something that customers want. And if you build agents with really high reliability standard, but are continuing pushing a level of abstraction, you then learn from your users how to get that next level of abstraction faster. So that's how you actually build the data flow. That's the critical path for the company. Everything we do is in service of that. So if you go zoom way, way back to Act One days, right? Like the core thing behind Act One is can we teach large model basically how to even actuate your computer? And I think we're one of the first places to have solved that and shown it and shown the generalization that you get when you give it various different workflows and texts. But I think from there on out, we really realized was that in order to get reliability, companies just do things in various different ways. You actually want these models to be able to get a lot better at having some specification of some guardrails for what it actually should be doing. And I think in conjunction with that, a giant thing that was really necessary is really fast multimodal models that are really good at understanding knowledge work and really good at understanding screens. And that is needs to kind of be the base for some of these agents. Back then we had to do a ton of research basically on how do we actually make that possible? Well, first off, like back in forgot exactly one month to 23, like there were no multimodal models really that you could use for things like this. And so we pushed really hard on stuff like the Fuyu architecture. I think one big hangover primarily academic focus for multimodal models is most multimodal models are primarily trained on like natural images, cat and dog photos, stuff that's come out of the camera. Coco. Yeah, right. And the Coco is awesome. Like I love Coco. I love TY. Like it's really helped the field. Right. But like that's the build one thing. I actually think it's really clear today. Multimodal models are the default foundation model, right? It's just going to supplant LLMs. Like you just train a giant multimodal model. And so for that though, like where are they going to be the most useful? They're going to be most useful in knowledge work tasks. That's where the majority of economic value is going to be. It's not in cat and dogs. Right. And so if that's what it is, what do you need to train? I need to train on like charts, graphs, tables, invoices, PDFs, receipts, unstructured data, UIs. That's just a totally different pre-training corpus. And so a depth spent a lot of time building that. And so the public for use and stuff aren't trained on our actual corpus, it's trained on some other stuff. But you take a lot of that data and then you make it really fast and make it really good at things like dense OCR on screens. And then now you have the right like raw putty to go make a good agent. So that's kind of like some of the modeling side, we've kind of only announced some of that stuff. We haven't really announced much of the agent's work, but that if you put those together with the correct product form factor, and I think the product form factor also really matters. I think we're seeing, and you guys probably see this a little bit more than I do, but we're seeing like a little bit of a pushback against the tyranny of chatbots as form factor. And I think that the reason why the form factor matters is the form factor changes what data you collect in the human feedback loop. And so I think we've spent a lot of time doing full vertical integration of all these bits in order to get to where we are.Swyx [00:25:44]: Yeah. I'll plug Amelia Wattenberger's talk at our conference, where she gave a little bit of the thinking behind like what else exists other than chatbots that if you could delegate to reliable agents, you could do. I was kind of excited at Adept experiments or Adept workflows, I don't know what the official name for it is. I was like, okay, like this is something I can use, but it seems like it's just an experiment for now. It's not your product.David [00:26:06]: So you basically just use experiments as like a way to go push various ideas on the design side to some people and just be like, yeah, we'll play with it. Actually the experiments code base underpins the actual product, but it's just the code base itself is kind of like a skeleton for us to go deploy arbitrary cards on the side.Swyx [00:26:22]: Yeah.Alessio [00:26:23]: Makes sense. I was going to say, I would love to talk about the interaction layer. So you train a model to see UI, but then there's the question of how do you actually act on the UI? I think there was some rumors about open app building agents that are kind of like, they manage the end point. So the whole computer, you're more at the browser level. I read in one of your papers, you have like a different representation, kind of like you don't just take the dome and act on it. You do a lot more stuff. How do you think about the best way the models will interact with the software and like how the development of products is going to change with that in mind as more and more of the work is done by agents instead of people?David [00:26:58]: This is, there's so much surface area here and it's actually one of the things I'm really excited about. And it's funny because I've spent most of my time doing research stuff, but there's like a whole new ball game that I've been learning about and I find it really cool. So I would say the best analogy I have to why Adept is pursuing a path of being able to use your computer like a human, plus of course being able to call APIs and being able to call APIs is the easy part, like being able to use your computer like a human is a hard part. It's in the same way why people are excited about humanoid robotics, right? In a world where you had T equals infinity, right? You're probably going to have various different form factors that robots could just be in and like all the specialization. But the fact is that humans live in a human environment. So having a human robot lets you do things that humans do without changing everything along the way. It's the same thing for software, right? If you go itemize out the number of things you want to do on your computer for which every step has an API, those numbers of workflows add up pretty close to zero. And so then many points along the way, you need the ability to actually control your computer like a human. It also lets you learn from human usage of computers as a source of training data that you don't get if you have to somehow figure out how every particular step needs to be some particular custom private API thing. And so I think this is actually the most practical path. I think because it's the most practical path, I think a lot of success will come from going down this path. I kind of think about this early days of the agent interaction layer level is a little bit like, do you all remember Windows 3.1? Like those days? Okay, this might be, I might be, I might be too old for you guys on this. But back in the day, Windows 3.1, we had this transition period between pure command line, right? Being the default into this new world where the GUI is the default and then you drop into the command line for like programmer things, right? The old way was you booted your computer up, DOS booted, and then it would give you the C colon slash thing. And you typed Windows and you hit enter, and then you got put into Windows. And then the GUI kind of became a layer above the command line. The same thing is going to happen with agent interfaces is like today we'll be having the GUI is like the base layer. And then the agent just controls the current GUI layer plus APIs. And in the future, as more and more trust is built towards agents and more and more things can be done by agents, if more UIs for agents are actually generative in and of themselves, then that just becomes a standard interaction layer. And if that becomes a standard interaction layer, what changes for software is that a lot of software is going to be either systems or record or like certain customized workflow execution engines. And a lot of how you actually do stuff will be controlled at the agent layer.Alessio [00:29:19]: And you think the rabbit interface is more like it would like you're not actually seeing the app that the model interacts with. You're just saying, hey, I need to log this call on Salesforce. And you're never actually going on salesforce.com directly as the user. I can see that being a model.David [00:29:33]: I think I don't know enough about what using rabbit in real life will actually be like to comment on that particular thing. But I think the broader idea that, you know, you have a goal, right? The agent knows how to break your goal down into steps. The agent knows how to use the underlying software and systems or record to achieve that goal for you. The agent maybe presents you information in a custom way that's only relevant to your particular goal, all just really leads to a world where you don't really need to ever interface with the apps underneath unless you're a power user for some niche thing.Swyx [00:30:03]: General question. So first of all, I think like the sort of input mode conversation. I wonder if you have any analogies that you like with self-driving, because I do think like there's a little bit of how the model should perceive the world. And you know, the primary split in self-driving is LiDAR versus camera. And I feel like most agent companies that I'm tracking are all moving towards camera approach, which is like the multimodal approach, you know, multimodal vision, very heavy vision, all the Fuyu stuff that you're doing. You're focusing on that, including charts and tables. And do you find that inspiration there from like the self-driving world? That's a good question.David [00:30:37]: I think sometimes the most useful inspiration I've found from self-driving is the levels analogy. I think that's awesome. But I think that our number one goal is for agents not to look like self-driving. We want to minimize the chances that agents are sort of a thing that you just have to bang your head at for a long time to get to like two discontinuous milestones, which is basically what's happened in self-driving. We want to be living in a world where you have the data flywheel immediately, and that takes you all the way up to the top. But similarly, I mean, compared to self-driving, like two things that people really undervalue is like really easy to driving a car down highway 101 in a sunny day demo. That actually doesn't prove anything anymore. And I think the second thing is that as a non-self-driving expert, I think one of the things that we believe really strongly is that everyone undervalues the importance of really good sensors and actuators. And actually a lot of what's helped us get a lot of reliability is a really strong focus on actually why does the model not do this thing? And the non-trivial amount of time, the time the model doesn't actually do the thing is because if you're a wizard of ozzing it yourself, or if you have unreliable actuators, you can't do the thing. And so we've had to fix a lot of those problems.Swyx [00:31:43]: I was slightly surprised just because I do generally consider the way most that we see all around San Francisco as the most, I guess, real case of agents that we have in very material ways.David [00:31:55]: Oh, that's absolutely true. I think they've done an awesome job, but it has taken a long time for self-driving to mature from when it entered the consciousness and the driving down 101 on a sunny day moment happened to now. Right. So I want to see that more compressed.Swyx [00:32:07]: And I mean, you know, cruise, you know, RIP. And then one more thing on just like, just going back on this reliability thing, something I have been holding in my head that I'm curious to get your commentary on is I think there's a trade-off between reliability and generality, or I want to broaden reliability into just general like sort of production readiness and enterprise readiness scale. Because you have reliability, you also have cost, you have speed, speed is a huge emphasis for a debt. The tendency or the temptation is to reduce generality to improve reliability and to improve cost, improve speed. Do you perceive a trade-off? Do you have any insights that solve those trade-offs for you guys?David [00:32:42]: There's definitely a trade-off. If you're at the Pareto frontier, I think a lot of folks aren't actually at the Pareto frontier. I think the way you get there is basically how do you frame the fundamental agent problem in a way that just continues to benefit from data? I think one of the main ways of being able to solve that particular trade-off is you basically just want to formulate the problem such that every particular use case just looks like you collecting more data to go make that use case possible. I think that's how you really solve. Then you get into the other problems like, okay, are you overfitting on these end use cases? You're not doing a thing where you're being super prescriptive for the end steps that the model can only do, for example.Swyx [00:33:17]: Then the question becomes, do you have one house model that you can then customize for each customer and you're fine-tuning them on each customer's specific use case?David [00:33:25]: Yeah.Swyx [00:33:26]: We're not sharing that. You're not sharing that. It's tempting, but that doesn't look like AGI to me. You know what I mean? That is just you have a good base model and then you fine-tune it.David [00:33:35]: For what it's worth, I think there's two paths to a lot more capability coming out of the models that we all are training these days. I think one path is you figure out how to spend, compute, and turn it into data. In that path, I consider search, RL, all the things that we all love in this era as part of that path, like self-play, all that stuff. The second path is how do you get super competent, high intelligence demonstrations from humans? I think the right way to move forward is you kind of want to combine the two. The first one gives you maximum sample efficiency for a little second, but I think that it's going to be hard to be running at max speed towards AGI without actually solving a bit of both.Swyx [00:34:16]: You haven't talked much about synthetic data, as far as I can tell. Probably this is a bit too much of a trend right now, but any insights on using synthetic data to augment the expensive human data?David [00:34:26]: The best part about framing AGI as being able to help people do things on computers is you have an environment.Swyx [00:34:31]: Yes. So you can simulate all of it.David [00:34:35]: You can do a lot of stuff when you have an environment.Alessio [00:34:37]: We were having dinner for our one-year anniversary. Congrats. Yeah. Thank you. Raza from HumanLoop was there, and we mentioned you were coming on the pod. This is our first-Swyx [00:34:45]: So he submitted a question.Alessio [00:34:46]: Yeah, this is our first, I guess, like mailbag question. He asked, when you started GPD 4 Data and Exist, now you have a GPD 4 vision and help you building a lot of those things. How do you think about the things that are unique to you as Adept, and like going back to like the maybe research direction that you want to take the team and what you want people to come work on at Adept, versus what is maybe now become commoditized that you didn't expect everybody would have access to?David [00:35:11]: Yeah, that's a really good question. I think implicit in that question, and I wish he were tier two so he can push back on my assumption about his question, but I think implicit in that question is calculus of where does advantage accrue in the overall ML stack. And maybe part of the assumption is that advantage accrues solely to base model scaling. But I actually believe pretty strongly that the way that you really win is that you have to go build an agent stack that is much more than that of the base model itself. And so I think like that is always going to be a giant advantage of vertical integration. I think like it lets us do things like have a really, really fast base model, is really good at agent things, but is bad at cat and dog photos. It's pretty good at cat and dog photos. It's not like soda at cat and dog photos, right? So like we're allocating our capacity wisely, right? That's like one thing that you really get to do. I also think that the other thing that is pretty important now in the broader foundation modeling space is I feel despite any potential concerns about how good is agents as like a startup area, right? Like we were talking about earlier, I feel super good that we're doing foundation models in service of agents and all of the reward within Adept is flowing from can we make a better agent? Because right now I think we all see that, you know, if you're training on publicly available web data, you put in the flops and you do reasonable things, then you get decent results. And if you just double the amount of compute, then you get predictably better results. And so I think pure play foundation model companies are just going to be pinched by how good the next couple of llamas are going to be and the next what good open source thing. And then seeing the really big players put ridiculous amounts of compute behind just training these base foundation models, I think is going to commoditize a lot of the regular LLMs and soon regular multimodal models. So I feel really good that we're just focused on agents.Swyx [00:36:56]: So you don't consider yourself a pure play foundation model company?David [00:36:59]: No, because if we were a pure play foundation model company, we would be training general foundation models that do summarization and all this other...Swyx [00:37:06]: You're dedicated towards the agent. Yeah.David [00:37:09]: And our business is an agent business. We're not here to sell you tokens, right? And I think like selling tokens, unless there's like a...Swyx [00:37:14]: Not here to sell you tokens. I love it.David [00:37:16]: It's like if you have a particular area of specialty, right? Then you won't get caught in the fact that everyone's just scaling to ridiculous levels of compute. But if you don't have a specialty, I find that, I think it's going to be a little tougher.Swyx [00:37:27]: Interesting. Are you interested in robotics at all? Just a...David [00:37:30]: I'm personally fascinated by robotics. I've always loved robotics.Swyx [00:37:33]: Embodied agents as a business, you know, Figure is like a big, also sort of open AI affiliated company that raises a lot of money.David [00:37:39]: I think it's cool. I think, I mean, I don't know exactly what they're doing, but...Swyx [00:37:44]: Robots. Yeah.David [00:37:46]: Well, I mean, that's a...Swyx [00:37:47]: Yeah. What question would you ask? If we had them on, what would you ask them?David [00:37:50]: Oh, I just want to understand what their overall strategy is going to be between now and when there's reliable stuff to be deployed. But honestly, I just don't know enough about it.Swyx [00:37:57]: And if I told you, hey, fire your entire warehouse workforce and, you know, put robots in there, isn't that a strategy? Oh yeah.David [00:38:04]: Yeah. Sorry. I'm not questioning whether they're doing smart things. I genuinely don't know what they're doing as much, but I think there's two things. One, I'm so excited for someone to train a foundation model of robots. It's just, I think it's just going to work. Like I will die on this hill, but I mean, like again, this whole time, like we've been on this podcast, we're just going to continually saying these models are basically behavioral cloners. Right. So let's go behavioral clone all this like robot behavior. Right. And then you figure out everything else you have to do in order to teach you how to solve a new problem. That's going to work. I'm super stoked for that. I think unlike what we're doing with helping humans with knowledge work, it just sounds like a more zero sum job replacement play. Right. And I'm personally less excited about that.Alessio [00:38:46]: We had a Ken June from InBoo on the podcast. We asked her why people should go work there and not at Adept.Swyx [00:38:52]: Oh, that's so funny.Alessio [00:38:54]: Well, she said, you know, there's space for everybody in this market. We're all doing interesting work. And she said, they're really excited about building an operating system for agent. And for her, the biggest research thing was like getting models, better reasoning and planning for these agents. The reverse question to you, you know, why should people be excited to come work at Adept instead of InBoo? And maybe what are like the core research questions that people should be passionate about to have fun at Adept? Yeah.David [00:39:22]: First off, I think that I'm sure you guys believe this too. The AI space to the extent there's an AI space and the AI agent space are both exactly as she likely said, I think colossal opportunities and people are just going to end up winning in different areas and a lot of companies are going to do well. So I really don't feel that zero something at all. I would say to like change the zero sum framing is why should you be at Adept? I think there's two huge reasons to be at Adept. I think one of them is everything we do is in the service of like useful agents. We're not a research lab. We do a lot of research in service of that goal, but we don't think about ourselves as like a classic research lab at all. And I think the second reason I work at Adept is if you believe that actually having customers and a reward signal from customers lets you build a GI faster, which we really believe, then you should come here. And I think the examples for why that's true is for example, our evaluations, they're not academic evals. They're not simulator evals. They're like, okay, we have a customer that really needs us to do these particular things. We can do some of them. These are the ones they want us to, we can't do them at all. We've turned those into evals, solve it, right? I think that's really cool. Like everybody knows a lot of these evals are like pretty saturated and the new ones that even are not saturated. You look at someone and you're like, is this actually useful? Right? I think that's a degree of practicality that really helps. Like we're equally excited about the same problems around reasoning and planning and generalization and all of this stuff. They're very grounded in actual needs right now, which is really cool.Swyx [00:40:45]: Yeah. This has been a wonderful dive. You know, I wish we had more time, but I would just leave it kind of open to you. I think you have broad thoughts, you know, just about

久违的一对一访谈回来啦！这次的嘉宾绝对重磅，贾扬清老师，关注AI领域的同学应该都听过他的鼎鼎大名！他在 UC Berkeley 博士期间创立了深度学习框架 Caffe, 很快成为行业事实标准。先后在 Google Brain, Facebook AI 从事最前沿的AI研究，随后又担任了阿里巴巴技术副总裁，领导大数据计算平台。2023年开始新征程，在硅谷创立了 Lepton AI. Hello World, who is OnBoard!? 作为AI和infra行业的行业领军人物，扬清老师是如何思考自己AI创业的方向的？他如何理解未来AI对于基础设施的需求，跟云计算这么多年的发展有哪些异同的地方？这一年以来，回到世界AI创新中心的硅谷，他对于AI和创业的理解、开发者工具和应用的价值、开源和闭源模型等等话题，都有怎样的思考迭代？ 我们不知不觉又聊了近两个小时，真是干货满满，你也能感受到扬清条理清晰、观点犀利，又温和儒雅，实在是太令人享受的谈话了。这大概就是播客的魅力，让我们在文字之外，感受到更真实鲜活的人。嘉宾长期在美国工作生活，有英文在所难免，不接受抱怨！Enjoy! 嘉宾介绍 贾扬清（推特：@jiayq），Lepton.ai 创始人。本科和研究生阶段就读于清华大学自动化专业，后赴加州大学伯克利分校攻读计算机科学博士。他在博士期间创立并开源了如今业内耳熟能详的深度学习框架Caffe，被微软、雅虎、英伟达、Adobe 等公司采用。2013年毕业后，他加入谷歌，是谷歌大脑 TensorFlow 的作者之一。2016年2月加盟Facebook，并开发出Caffe2Go、Caffe2、PyTorch等深度学习框架。2019 年加入阿里巴巴，担任阿里巴巴集团副总裁、阿里云智能计算平台事业部总裁。 嘉宾主持：戴雨森，真格基金合伙人，清华大学工业工程系2004级校友，曾在斯坦福大学管理科学与工程系就读。戴雨森22岁时参与创办了知名互联网上市公司聚美优品，主管互联网产品、运营、市场投放、品类等。加入真格基金之后，主要关注人工智能方向投资。 OnBoard! 主持：Monica：美元VC投资人，前 AWS 硅谷团队+ AI 创业公司打工人，公众号M小姐研习录 (ID: MissMStudy) 主理人 | 即刻：莫妮卡同学 我们都聊了什么 02:14 主持和嘉宾的自我介绍，Lepton 最近一篇论文为什么值得关注？ 06:00 Lepton AI是做什么的，为什么称之为 AI cloud company? 10:02 为什么想要成立 Lepton AI? 11:50 设计针对AI的基础设施难点在哪里？跟传统云厂商和HPC的差别是什么？ 19:46 为什么说现在我们不需要担心AI推理成本？未来提升的空间有多少？硬件和软件还可能有哪些突破？ 25:27 开发者如何选择AI基础设施和响应的开发工具？为什么 leaderboard 是不够的？ 28:49 Nvidia 会有新的挑战者吗？什么是“不可能三角”？ 33:48 MLOps 是个伪命题？！AI 需要的开发工具是怎样的？ 39:01 应用开发门槛越来越低，如何思考AI应用的价值？微软20年前的海报给了我们怎样的启发？ 44:47 AI native 的组织是怎样的？ 54:51 开源和闭源、专用和通用模型未来的关系？未来会 one model rules all 吗？ 64:24 创业之后有什么感受和收获？去年年初提出的“三个基本假设”，这一年有什么变化？ 67:56 未来AI应用和平台的市场格局会发生怎样的变化？ 70:01 为什么说我们低估了颠覆的难度？期待5年后AI可以完成什么？ 76:59 快问快答：喜欢的AI产品，推荐的书籍，解压的方式，想要问 AI 什么问题？ 我们提到的内容 DistriFusion: Distributed Parallel Inference for High-Resolution Diffusion Models, Paper, Code Meta research: Training ImageNet in 1 Hour AI inference leaderboard Lepton Search, Code Perplexity 推荐的书：菊与刀 参考文章 贾扬清的个人网站 贾扬清：三个基础假设 贾扬清：ChatGPT，和聪明地设计 Infra Twitter 讨论：Are LLM APIs losing money? Does One Large Model Rule Them All? 欢迎关注M小姐的微信公众号，了解更多中美软件、AI与创业投资的干货内容！M小姐研习录 (ID: MissMStudy) 欢迎在评论区留下你的思考，与听友们互动。喜欢 OnBoard! 的话，也可以点击打赏，请我们喝一杯咖啡！如果你用 Apple Podcasts 收听，也请给我们一个五星好评，这对我们非常重要。 OnBoard! 终于成立听友群啦！新年新气象，加入Onboard听友群，结识到高质量的听友们，我们还会组织线下主题聚会，开放实时旁听播客录制，嘉宾互动等新的尝试。添加任意一位小助手微信，onboard666, 或者 Nine_tunes, 发送你的姓名、公司和职位，小助手会拉你进群。期待你来！

In this conversation, we have an illuminating discussion with Waymo co-CEO Dmitri Dolgov moderated by Sebastian Thrun who started the Google Self-Driving Car project back in the day, and is also the founder of GoogleX, Google Brain, Waymo and Udacity. Tune in as they dive into the story behind Waymo's fully autonomous cars, and the road that lies ahead in this fascinating industry that is changing the way we move in the world.

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Why I think it's net harmful to do technical safety research at AGI labs, published by Remmelt on February 7, 2024 on LessWrong. IMO it is harmful on expectation for a technical safety researcher to work at DeepMind, OpenAI or Anthropic. Four reasons: Interactive complexity. The intractability of catching up - by trying to invent general methods for AI corporations to somehow safely contain model interactions, as other engineers scale models' combinatorial complexity and outside connectivity. Safety-capability entanglements Commercialisation. Model inspection and alignment techniques can support engineering and productisation of more generally useful automated systems. Infohazards. Researching capability risks within an AI lab can inspire researchers hearing about your findings to build new capabilities. Shifts under competitive pressure DeepMind merged with Google Brain to do commercialisable research, OpenAI set up a company and partnered with Microsoft to release ChatGPT, Anthropic pitched to investors they'd build a model 10 times more capable. If you are an employee at one of these corporations, higher-ups can instruct you to do R&D you never signed up to do.[1] You can abide, or get fired. Working long hours surrounded by others paid like you are, by a for-profit corp, is bad for maintaining bearings and your epistemics on safety.[2] Safety-washing. Looking serious about 'safety' helps labs to recruit idealistic capability researchers, lobby politicians, and market to consumers. 'let's build AI to superalign AI' 'look, pretty visualisations of what's going on inside AI' This is my view. I would want people to engage with the different arguments, and think for themselves what ensures that future AI systems are actually safe. ^ I heard via via that Google managers are forcing DeepMind safety researchers to shift some of their hours to developing Gemini for product-ready launch. I cannot confirm whether that's correct. ^ For example, I was in contact with a safety researcher at an AGI lab who kindly offered to read my comprehensive outline on the AGI control problem, to consider whether to share with colleagues. They also said they're low energy. They suggested I'd remind them later, and I did, but they never got back to me. They're simply too busy it seems. Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org

Read the full transcript here. What does it really mean to align an AI system with human values? What would a powerful AI need to do in order to do "what we want"? How does being an assistant differ from being an agent? Could inter-AI debate work as an alignment strategy, or would it just result in arguments designed to manipulate humans via their cognitive and emotional biases? How can we make sure that all human values are learned by AIs, not just the values of humans in WEIRD societies? Are our current state-of-the-art LLMs politically left-leaning? How can alignment strategies take into account the fact that our individual and collective values occasionally change over time?Geoffrey Irving is an AI safety researcher at DeepMind. Before that, he led the Reflection Team at OpenAI, was involved in neural network theorem proving at Google Brain, cofounded Eddy Systems to autocorrect code as you type, and worked on computational physics and geometry at Otherlab, D. E. Shaw Research, Pixar, and Weta Digital. He has screen credits on Ratatouille, WALL•E, Up, and Tintin. Learn more about him at his website, naml.us.Further reading:Gandalf: An Educational Game Demonstrating Security Vulnerabilities in Large Language Models"AI safety via debate""Claude's Constitution" Staff Spencer Greenberg — Host / Director Josh Castle — Producer Ryan Kessler — Audio Engineer Uri Bram — Factotum WeAmplify — Transcriptionists Miles Kestran — Marketing Music Lee Rosevere Josh Woodward Broke for Free zapsplat.com wowamusic Quiet Music for Tiny Robots Affiliates Clearer Thinking GuidedTrack Mind Ease Positly UpLift [Read more]

Jonathan Frankle works as Chief Scientist (Neural Networks) at MosaicML (recently acquired by Databricks), a startup dedicated to making it easy and cost-effective for anyone to train large-scale, state-of-the-art neural networks. He leads the research team. MLOps podcast #205 with Jonathan Frankle, Chief Scientist (Neural Networks) at Databricks, The Myth of AI Breakthroughs, co-hosted by Denny Lee, brought to us by our Premium Brand Partner, Databricks. // Abstract Jonathan takes us behind the scenes of the rigorous work they undertake to test new knowledge in AI and to create effective and efficient model training tools. With a knack for cutting through the hype, Jonathan focuses on the realities and usefulness of AI and its application. We delve into issues such as face recognition systems, the 'lottery ticket hypothesis,' and robust decision-making protocols for training models. Our discussion extends into Jonathan's interesting move into the world of law as an adjunct professor, the need for healthy scientific discourse, his experience with GPUs, and the amusing claim of a revolutionary algorithm called Qstar. // Bio Jonathan Frankle is Chief Scientist (Neural Networks) at Databricks, where he leads the research team toward the goal of developing more efficient algorithms for training neural networks. He arrived via Databricks' $1.3B acquisition of MosaicML as part of the founding team. He recently completed his PhD at MIT, where he empirically studied deep learning with Prof. Michael Carbin, specifically the properties of sparse networks that allow them to train effectively (his "Lottery Ticket Hypothesis" - ICLR 2019 Best Paper). In addition to his technical work, he is actively involved in policymaking around challenges related to machine learning. He earned his BSE and MSE in computer science at Princeton and has previously spent time at Google Brain and Facebook AI Research as an intern and Georgetown Law as an Adjunct Professor of Law. // MLOps Jobs board https://mlops.pallet.xyz/jobs // MLOps Swag/Merch https://mlops-community.myshopify.com/ // Related Links Website: www.jfrankle.com Facial recognition: perpetuallineup.orgThe Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networksby Jonathan Frankle and Michael Carbin paper: https://arxiv.org/abs/1803.03635 --------------- ✌️Connect With Us ✌️ ------------- Join our slack community: https://go.mlops.community/slack Follow us on Twitter: @mlopscommunity Sign up for the next meetup: https://go.mlops.community/register Catch all episodes, blogs, newsletters, and more: https://mlops.community/ Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/ Connect with Denny on LinkedIn: https://linkedin.com/in/dennyglee Connect with Jonathan on LinkedIn: https://www.linkedin.com/in/jfrankle/ Timestamps: [00:00] Jonathan's preferred coffee [01:16] Takeaways [07:19] LM Avalanche Panel Surprise [10:07] Adjunct Professor of Law [12:59] Low facial recognition accuracy [14:22] Automated decision making human in the loop argument [16:09] Control vs. Outsourcing Concerns [18:02] perpetuallineup.org [23:41] Face Recognition Challenges [26:18] The lottery ticket hypothesis [29:20] Mosaic Role: Model Expertise [31:40] Expertise Integration in Training [38:19] SLURM opinions [41:30] GPU Affinity [45:04] Breakthroughs with QStar [49:52] Deciphering the noise advice [53:07] Real Conversations [55:47] How to cut through the noise [1:00:12] Research Iterations and Timelines [1:02:30] User Interests, Model Limits [1:06:18] Debugability [1:08:00] Wrap up

Как подготовиться к олимпиадам так, чтобы выиграть золото на международном этапе? Насколько красивой могут быть алгебра с геометрией? Кто такой Quantative Researcher и что нужно изучать, чтобы им стать? Сегодня у нас праздник для математиков и для всех, кто считает, что математика — царица наук. У нас в гостях Игорь Ганичев, легенда КТЛ, победитель международной олимпиады по математике, человек, прошедший школу MIT, CU Berkeley, VMware, Google Brain и исследователь в частном хедж-фонде. С Игорем мы успели поговорить про интересные курсы из MIT и Berkeley, интересные методики для анализа цены акций и ИИ-гонку. А ещё попросили советов по тому, как обучать детей программированию и как эффективно преподавать математику. Наслаждайтесь выпуском, любите математику и смотрите nFactorial Podcast, друзья! Благодарим международный технопарк Astana Hub за поддержку данного эпизода! https://astanahub.com

This week we talk about regulatory capture, Open AI, and Biden's executive order.We also discuss the UK's AI safety summit, open source AI models, and flogging fear. Recommended Book: The Resisters by Gish JenTranscriptRegulatory capture refers to the corruption of a regulatory body by entities to which the regulations that body creates and enforces, apply.So an organization that wants to see less funding for public schools and more for private and home schooling options getting one of their people into a position at the Department of Education, or someone from Goldman Sachs or another, similar financial institution getting shoehorned into a position at the Federal Reserve, could—through some lenses at least, and depending on how many connections those people in those positions have to those other, affiliated, ideological and commercial institutions—could be construed as engaging in regulatory capture, because they're now able to control the levers of regulation that apply to their own business or industry, or their peers, the folks they previously worked with and people to whom they maybe owe favors, or vice versa, and that could lead to regulations that are more favorable to them and their preferred causes, and those of their fellow travelers.This is in contrast to regulatory bodies that apply limits to such businesses and organizations, figuring out where they might overstep or lock in their own power at the expense of the industry in which they operate, and slowly, over time, plugging loopholes, finding instances of not-quite-illegal misdeeds that nonetheless lead to negative outcomes, and generally being the entity in charge in spaces that might otherwise be dominated by just one or two businesses that can kill off all their competition and make things worse for consumers and workers.Often, rather than regulatory capture being a matter of one person from a group insinuating themselves into the relevant regulatory body, the regulatory body, itself, will ask representatives from the industry they regulate to help them make law, because, ostensibly at least, those regulatees should know the business better than anyone else, and in helping to create their own constraints—again, ostensibly—they should be more willing to play by the rules, because they helped develop the rules to which they're meant to abide, and probably helped develop rules that they can live with and thrive under; because most regulators aren't trying to kill ambition or innovation or profit, they're just trying to prevent abuses and monopolistic hoarding.This sort of capture has taken many shapes over the years, and occurred at many scales.In the late-19th century, for instance, railroad tycoons petitioned the US government for regulation to help them bypass a clutter of state-level regulations that were making it difficult and expensive for them to do business, and in doing so—in asking to be regulated and helping the federal government develop the applicable regulations—they were able to make their own lives easier, while also creating what was effectively a cartel for themselves with the blessing of the government that regulated their power; the industry as it existed when those regulations were signed into law, was basically locked into place, in such a way that no new competitors could practically arise.Similar efforts have been launched, at times quite successfully, by entities in the energy space, across various aspects of the financial world, and in just about every other industry you can imagine, from motorcyclists' protective clothing to cheerleading competitions to aviation and its many facets—all have been to some degree and at some point allegedly regulatorily captured so that those being regulated to some degree control the regulations under which they operate, and which as a consequence has at times allowed them to create constraints that benefit them and entrench their own power, rather than opening their industry up and increasing competition, safety, and the treatment and benefits afforded to customers and workers, as is generally the intended outcome of these regulations.What I'd like to talk about today is the burgeoning world of artificial intelligence and why some players in this space are being accused of attempting the time-tested tactic of regulatory capture at a pivotal moment of AI development and deployment.—At the tail-end of October, 2023, US President Biden announced that he was signing a fairly expansive executive order on AI: the first of its kind, and reportedly the first step toward still-greater and more concrete regulation.A poll conducted by the AI Policy Institute suggests that Americans are generally in favor of this sort of regulatory move, weighing in at 68% in favor of the initiative, which is a really solid in-favor number, especially at a moment as politically divided as this one, and most of the companies working in this space—at least at a large enough scale to show up on the map for AI at this point—seem to be in favor of this executive order, as well, with some caveats that I'll get to in a bit.That indicates the government probably got things pretty close to where they need to be, in terms of folks actually adhering to these rules, though it's important to note that part of why there's such broad acceptance of the tenets of this order is that there aren't any real teeth to these rules: it's largely voluntary stuff, and mostly only applies to the anticipated next generation of AI—the current generation isn't powerful enough to fall under its auspices, in most cases, so AI companies don't need to do much of anything yet to adhere to these standards, and when they eventually do need to do something to remain in accordance with them, it'll mostly be providing reports to government employees so they can keep tabs on developments, including those happening behind close doors, in this space.Now that is not nothing: at the moment, this industry is essentially a black box as far as would-be regulators are concerned, so simply providing a process by which companies working on advanced AI and AI applications can keep the government informed on their efforts is a big step that raises visibility from 0 to some meaningful level.It also provides mechanisms through which such entities can get funding from the government, and pathways through which international AI experts can come to the United States with less friction than would be the case for folks without that expertise.So AI industry entities generally like all this because it's easy for them to work with, is flexible enough not to punish them if they fail in some regard, but it also provides them with more resources, both monetary and human, and sets the US up, in many ways, to maintain its current purported AI dominance well into the future, despite essentially everyone—especially but not exclusively China—investing a whole lot to catch up and surpass the US in the coming years.Another response to this order, though, and the regulatory infrastructure it creates, was voiced by the founder of Google Brain, Andrew Ng, who has been working on AI systems and applications for a long time, and who basically says that some of the biggest players in AI, today, are playing up the idea that artificial intelligence systems might be dangerous, even to the point of being world-ending, because they hope to create exactly this kind of regulatory framework at this exact moment, because right now they are the kings of the AI ecosystem, and they're hoping to lock that influence in, denying easy access to any future competitors.This theory is predicated on that concept I mentioned in the intro, regulatory capture, and history is rich with examples of folks in positions of power in various spaces telling their governments to put their industry on lockdown, and making the case for why this is necessary, because they know, in doing so, their position at the top will probably be locked in, because it will become more difficult and expensive and thus, out of reach, for any newer, smaller, not already influential and powerful competitor, to then challenge them moving forward.One way this might manifest in the AI space, according to Ng, is through the licensing of powerful AI models—essentially saying if you want to use the more powerful AI systems for your product or research, you need to register with the government, and you need to buy access, basically, from one of these government-sanctioned providers. Only then will we allow you to play in this potentially dangerous space with these highest-end AI models.This, in turn, would substantially reduce innovation, as other entities wouldn't be able to legally evolve their AI in different directions, at least not at a high level, and it would make today's behemoths—the OpenAI's and Meta's of the world—all but invulnerable to future challenges, because their models would be the ones made available to everyone else to use; no one else could compete, not practically, at least.This would be not-great for smaller, upstart AI companies, but it would be especially detrimental to open source large language models—versions of the most popular, LLM-based AI systems that're open to the public to mess around with and use however they see fit, rather than being controlled and sold by a single company.These models would be unlikely to have the resources or governing body necessary to step into the position of regulator-approved moderator of potentially dangerous AI systems, and the open source credo doesn't really play well with that kind of setup to begin with, as the idea is that all the code is open and available to take and use and change, so locking it down at all would violate those principles; and this sort of regulatory approach would be all about the lockdown, on fears of bad actors getting their hands on high-end AI systems—fears that have been flogged by entities like OpenAI.So that collection of fears are potentially fueling the relatively fast-moving regulatory developments related to AI in the US, right now; regulation, by the way, that's typically slower-moving in the US, which is part of why this is so notable.This is not a US-exclusive concern, though, nor is this executive order the only big, new regulatory effort in this space.At a summit in the UK just days after the US executive order was announced, AI companies from around the world, and those who govern such entities, met up to discuss the potential national security risks inherent in artificial intelligence tools, and to sign a legally non-binding agreement to let their governments test their newest, most powerful models for risks before they're released to the public.The US participated in this summit, as well, and a lot of these new rules overlap with each other, as the executive order shares a lot of tenets with the agreement signed at that meeting in the UK—though the EO was US-specific and included non-security elements, as well, and that will be the case for laws and orders passed in the many different countries to which these sorts of global concerns apply, each with their own approach to implementing those more broadly agreed-upon specifics at the national level.This summit announced the creation of a international panel of experts who will publish an annual report on the state of the art within the AI space, especially as it applies to national security risks, like misinformation and cybersecurity issues, and when questioned about whether the UK should take things a step further, locking some of these ideas and rules into place and making them legal requirements rather than things corporations agree to do but aren't punished for not doing, the Prime Minister, Rishi Sunak said, in essence, that this sort of thing takes time; and that's a sentiment that's been echoed by many other lawmakers and by people within this industry, as well.We know there need to be stricter and more enforceable regulations in this space, but because of where we are with this collection of technologies and the culture and rules and applications of them, right now, we don't really know what laws would make the most sense, in other words.No nation wants to tie its own hands in developing increasingly useful and powerful AI tools, and moving too fast on the concrete versions of these sort of agreements could end up doing exactly that; there's no way to know what the best rules and regulations will be, yet, because we're standing at the precipice of what looks like a long journey toward a bunch of new discoveries and applications.That's why the US executive order is set up the way it is, too: Biden and his advisors don't want to slow down the development in this space within the US, they want to amplify it, while also providing some foundational structure for whatever they decide needs to be built next—but those next-step decisions will be shaped by how these technologies and industries evolve over the next few years.The US and other countries are also setting up agencies and institutes and all sorts of safety precautions related to this space, but most of them lack substance at this point, and as with the aforementioned regulations, these agency setups are primarily just first draft guide rails, if that, at this point.Notably, the EU seems to be orienting around somewhat sterner regulations, but they haven't been able to agree on anything concrete quite yet, so despite typically taking the lead on this sort of thing, the US is a little bit ahead of the EU in terms of AI regulation right now—though it's likely that when the EU does finally put something into place, it'll be harder-core than what the US has, currently.A few analysts in this space have argued that these new regulations—lightweight as they are, both on the global and US level—by definition will hobble innovation because regulations tend to do that: they're opinionated about what's important and what's not, and that then shapes the direction makers in the regulated space will tend go.There's also a chance that, as I mentioned before, that this set of regulations laid out in this way, will lock the power of incumbent AI companies into place, protecting them from future competitors, and in doing so also killing off a lot of the forces of innovation that would otherwise lead to unpredictable sorts of outcomes.One big question, then, is how light a touch these initial regulations will actually end up having, how the AI and adjacent industries will reshape themselves to account for these and predicted future regulations, and to what degree open source alternatives—and other third-party alternatives, beyond the current incumbents—will be able to step in and take market share, nudging things in different directions, and potentially either then being incorporated into and shaping those future, more toothy regulations, or halting the deployment of those regulations by showing that the current direction of regulatory development no longer makes sense.We'll also see how burdensome the testing and other security-related requirements in these initial rules end up being, as there's a chance more attention and resources will shift toward lighter-weight, less technically powerful, but more useful and deployable versions of these current AI tools, which is already something that many entities are experimenting with, because that comes with other benefits, like being able to run AI on devices like a smartphone, without needing to connect, through the internet, to a huge server somewhere.Refocusing on smaller models could also allow some developers and companies to move a lot faster than their more powerful but plodding and regulatorily hobbled kin, rewiring the industry in their favor, rather than toward those who are currently expected to dominate this space for the foreseeable future.Show NotesOn the EOhttps://www.aijobstracker.com/ai-executive-orderReactions to EOhttps://archive.ph/RdpLhhttps://theaipi.org/poll-biden-ai-executive-order-10-30/https://www.nytimes.com/2023/10/30/us/politics/biden-ai-regulation.html?ref=readtangle.comhttps://qz.com/does-anyone-not-like-bidens-new-guidelines-on-ai-1850974346https://archive.ph/wwRXjhttps://www.afr.com/technology/google-brain-founder-says-big-tech-is-lying-about-ai-human-extinction-danger-20231027-p5efnzhttps://twitter.com/ylecun/status/1718670073391378694?utm_source=substack&utm_medium=emailhttps://stratechery.com/2023/attenuating-innovation-ai/First take on EOWhat EO means for openness in AIBiden's regulation planshttps://www.reuters.com/technology/eu-lawmakers-face-struggle-reach-agreement-ai-rules-sources-2023-10-23/https://archive.ph/IwLZuhttps://techcrunch.com/2023/11/01/politicians-commit-to-collaborate-to-tackle-ai-safety-us-launches-safety-institute/https://indianexpress.com/article/explained/explained-sci-tech/on-ai-regulation-the-us-steals-a-march-over-europe-amid-the-uks-showpiece-summit-9015032/ This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit letsknowthings.substack.com/subscribe

This is a recap of the top 10 posts on Hacker News on October 30th, 2023.This podcast was generated by wondercraft.ai(00:37): I accidentally saved my company half a million dollarsOriginal post: https://news.ycombinator.com/item?id=38069710&utm_source=wondercraft_ai(02:25): Apple unveils M3, M3 Pro, and M3 Max, the most advanced chips for a PCOriginal post: https://news.ycombinator.com/item?id=38078063&utm_source=wondercraft_ai(04:20): Gmail, Yahoo announce new 2024 authentication requirements for bulk sendersOriginal post: https://news.ycombinator.com/item?id=38074992&utm_source=wondercraft_ai(06:01): Use YouTube to improve your English pronunciationOriginal post: https://news.ycombinator.com/item?id=38074701&utm_source=wondercraft_ai(07:40): Private equity is devouring the U.S. economyOriginal post: https://news.ycombinator.com/item?id=38069197&utm_source=wondercraft_ai(09:42): The costs of microservices (2020)Original post: https://news.ycombinator.com/item?id=38069915&utm_source=wondercraft_ai(11:47): Google Brain founder says big tech is lying about AI dangerOriginal post: https://news.ycombinator.com/item?id=38072218&utm_source=wondercraft_ai(13:37): AI.govOriginal post: https://news.ycombinator.com/item?id=38067206&utm_source=wondercraft_ai(15:24): The Grug Brained Developer (2022)Original post: https://news.ycombinator.com/item?id=38076886&utm_source=wondercraft_ai(17:01): Global CO2 LevelsOriginal post: https://news.ycombinator.com/item?id=38072267&utm_source=wondercraft_aiThis is a third-party project, independent from HN and YC. Text and audio generated using AI, by wondercraft.ai. Create your own studio quality podcast with text as the only input in seconds at app.wondercraft.ai. Issues or feedback? We'd love to hear from you: team@wondercraft.ai

Todd breaks down four AI articles, leading with an article by Andrew Ng, the co-founder of Google Brain, lifting the veil on the strategies employed by Big Tech giants in the realm of artificial intelligence. He posits that some of these companies might be inflating the risks associated with AI. But what's the endgame? According … Continue reading Big Tech & AI: Power Play or Genuine Concern? #1702 → The post Big Tech & AI: Power Play or Genuine Concern? #1702 appeared first on Geek News Central.

One thing that some are betting on is that emotions and emotional intelligence is something that's very different than what everyone else is doing right now with AI.  We are excited to bring Ryan Benmalek on to talk about the rise of emotional AI and how machines are learning to understand and respond to human emotions.  Ryan is a co-founder and CEO of Daimon Labs, a partially remote, partially NYC-based team from the top AI research labs like Google Brain, Microsoft Research, and FAIR.  They are innovating a new type of language model that provides human-like empathy and companionship.   Waitlist: Join Blog: 1–7x Consumer GPU Scaling for Large Language Modeling Experiments LinkedIn: The Rise of Emotional AI: How Machines Are Learning to Understand and Respond to Human Emotions

Andrew Ng, PhD, a distinguished authority in the field of AI, is known for founding DeepLearning.AI and multiple other ventures. He also co-founded and led Google Brain and serves as an Adjunct Professor in Stanford University's Computer Science Department. In this episode, he is joined by Vijay Pande, founding partner of a16z Bio + Health.Andrew has thought deeply about the implications of integrating AI into many areas of our lives, going so far as to put out a public social media call for people who believe AI is dangerous to speak with him. He and Vijay discussed this, as well as how AI could become foundational to many industries — and what needs to happen to make that future a reality.

Want to help define the AI Engineer stack? Have opinions on the top tools, communities and builders? We're collaborating with friends at Amplify to launch the first State of AI Engineering survey! Please fill it out (and tell your friends)!If AI is so important, why is its software so bad?This was the motivating question for Chris Lattner as he reconnected with his product counterpart on Tensorflow, Tim Davis, and started working on a modular solution to the problem of sprawling, monolithic, fragmented platforms in AI development. They announced a $30m seed in 2022 and, following their successful double launch of Modular/Mojo

with @alive_eth @danboneh @smc90This week's all-new episode covers the convergence of two important, very top-of-mind trends: AI (artificial intelligence) & blockchains/ crypto. These domains together have major implications for how we all live our lives everyday; so this episode is for anyone just curious about, or already building in the space. The conversation covers topics ranging from deep fakes, bots, and the need for proof-of-humanity in a world of AI; to big data, large language models like ChatGPT, user control, governance, privacy and security, zero knowledge and zkML; to MEV, media, art, and much more. Our expert guests (in conversation with host Sonal Chokshi) include: Dan Boneh, Stanford Professor (and Senior Research Advisor at a16z crypto), a cryptographer who's been working on blockchains for over a decade and who specializes in cryptography, computer security, and machine learning -- all of which intersect in this episode;Ali Yahya, general partner at a16z crypto, who also previously worked at Google -- where he not only worked on a distributed system for a fleet of robots (a sort of "collective reinforcement learning") but also worked on Google Brain, where he was one of the core contributors to the machine learning library TensorFlow built at Google.The first half of the hallway-style conversation between Ali & Dan (who go back together as student and professor at Stanford) is all about how AI could benefit from crypto, and the second half on how crypto could benefit from AI... the thread throughout is the tension between centralization vs. decentralization.  So we also discuss where the intersection of crypto and AI can bring about things that aren't possible by either one of them alone...pieces referenced in this episode/ related reading:The Next Cyber Reasoning System for Cyber Security (2023) by Mohamed Ferrag, Ammar Battah, Norbert Tihanyi, Merouane Debbah, Thierry Lestable, Lucas CordeiroA New Era in Software Security: Towards Self-Healing Software via Large Language Models and Formal Verification (2023) by  Yiannis Charalambous, Norbert Tihanyi, Ridhi Jain, Youcheng Sun, Mohamed Ferrag, Lucas CordeiroFixing Hardware Security Bugs with Large Language Models (2023) by Baleegh Ahmad, Shailja Thakur, Benjamin Tan, Ramesh Karri, Hammond PearceDo Users Write More Insecure Code with AI Assistants? (2022) by Neil Perry, Megha Srivastava, Deepak Kumar, Dan BonehAsleep at the Keyboard? Assessing the Security of GitHub Copilot's Code Contributions (2021) by Hammond Pearce, Baleegh Ahmad, Benjamin Tan, Brendan Dolan-Gavitt, Ramesh KarriVoting, Security, and Governance in Blockchains (2019) with Ali Yahya and Phil Daian    As a reminder: none of the following should be taken as investment, legal, business, or tax advice; please see a16z.com/disclosures for more important information -- including to a link to a list of our investments – especially since we are investors in companies mentioned in this episode. Stay Updated: Find a16z on Twitter: https://twitter.com/a16zFind a16z on LinkedIn: https://www.linkedin.com/company/a16zSubscribe on your favorite podcast app: https://a16z.simplecast.com/Follow our host: https://twitter.com/stephsmithioPlease note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures.

Does the convergence of tech and biology hold the key to reshaping biotech's data infrastructure? Our recent chat with Alfredo Andere, Co-Founder and CEO at LatchBio, certainly supports this notion. This episode offers a deep dive into Alfredo's journey from being part of the Google Brain team to co-founding Latch Bio - a company that is making waves in the biotech industry with its innovative solutions and has raised a whopping $33 million.First In Human is a biotech-focused podcast that interviews industry leaders and investors to learn about their journey to in-human clinical trials. Presented by Vial, a tech-enabled CRO, hosted by Simon Burns, CEO & Co-Founder. Episodes launch weekly on Tuesdays. To view the full transcript of this episode, click here. Interested in being featured as a guest on First In Human? Please reach out to catie@vial.com.

Noam Shazeer is the co-founder and CEO of Character.AI, a full-stack AI computing platform that gives people access to their own flexible superintelligence. A renowned computer scientist and researcher, Shazeer is one of the foremost experts in artificial intelligence (AI) and natural language processing (NLP). He is a key author for the Transformer, a revolutionary deep learning model enabling language understanding, machine translation, and text generation that has become the foundation of many NLP models. A former member of the Google Brain team, Shazeer led the development of spelling corrector capabilities within Gmail, the algorithm at the heart of AdSense.   In Today's Episode with Noam Shazeer We Discuss: 1. Entry into the World of AI and NLP: How did Noam first make his way into the world of AI and come to work on spell corrector with Google? What are 1-2 of his biggest takeaways from spending 20 years at Google? What does Noam know now that he wishes he had known when he started Character? 2. Model Size or Data Size: What is more important, the size of the data or the size of the model? Does Noam agree that "we will not use models in a year that we have today?" What is the lifespan of a model? Does Noam agree that the companies that win are those that are able to switch between models with the most ease? With the majority of data being able to be downloaded from the internet, is there real value in data anymore? 3. The Biggest Barriers: What is the single biggest barrier to Character today? What are the most challenging elements of model training? Why did they need to spend $2M to train an early model? What are the most difficult elements of releasing a horizontal product with so many different use cases? Where does the value accrue in the race for AI dominance; startups or incumbents? 4. AI's Role on Society: Why does Noam believe that AI can create greater not worse human connections? Why is Noam not concerned by the speed of adoption of AI tools? What does Noam know about AI's impact on society that the world does not see?

Colin Murdoch is the chief business officer at Google DeepMind. He joins Big Technology Podcast for a conversation about artificial general intelligence, discussing why we want to get there at all, and what the path looks like. We also discuss DeepMind's merger with Google Brain, how pursuing the AI business changes Google, and how DeepMind's AlphaFold AI is revolutionizing the healthcare space. Tune in for a dynamic conversation with one of the world's leading AI executives. --- Enjoying Big Technology Podcast? Please rate us five stars ⭐⭐⭐⭐⭐ in your podcast app of choice. For weekly updates on the show, sign up for the pod newsletter on LinkedIn: https://www.linkedin.com/newsletters/6901970121829801984/ Questions? Feedback? Write to: bigtechnologypodcast@gmail.com

In this episode, Nathan sits down with Paige Bailey, Lead Product Manager of Generative Models at Google Deepmind. In this conversation, they discuss what it's like to be a PM for an LLM as opposed to an app, defining ideal LLM behaviour, and reasoning - how do you distinguish real abilities vs pattern matching? If you're looking for an ERP platform, check out our sponsor, NetSuite: http://netsuite.com/cognitive RECOMMENDED PODCAST: The HR industry is at a crossroads. What will it take to construct the next generation of incredible businesses – and where can people leaders have the most business impact? Hosts Nolan Church and Kelli Dragovich have been through it all, the highs and the lows – IPOs, layoffs, executive turnover, board meetings, culture changes, and more. With a lineup of industry vets and experts, Nolan and Kelli break down the nitty-gritty details, trade offs, and dynamics of constructing high performing companies. Through unfiltered conversations that can only happen between seasoned practitioners, Kelli and Nolan dive deep into the kind of leadership-level strategy that often happens behind closed doors. Check out the first episode with the architect of Netflix's culture deck Patty McCord. https://link.chtbl.com/hrheretics TIMESTAMPS: (00:00) Episode Preview (00:01:15) Introducing Paige Bailey (00:04:21) Paige's background at Google Brain and the Deepmind merger (00:07:00) PM for a LLM vs being a PM for an app (00:11:21) The development timeline and compute budget of PaLM-2 (00:14:30) Paige's role in the PaLM 2 project (00:15:30) Sponsors: Netsuite | Omneky (00:17:26) Defining desired capabilities for PaLM-2 (00:19:17) The amount of work that went into elevating PaLM 2 from PaLM 1 (00:20:28) Has Google lost its ability to ship? (00:24:240) Paige's "eureka" moment seeing GitHub Copilot capabilities (00:27:47) Competing PaLM 2 with other models (00:32:20) Grokking and the predictability of emergent capabilities (00:37:30) Citizen scientists and the multilingual capabilities of PaLM 2 (00:39:29) Distinguishing real reasoning vs pattern matching (00:45:51) Products using PaLM-2 that people should try (00:50:35) Most exciting AI projects that you can try out (00:52:29) Curriculum learning and successor to the transformer LINKS: PaLM 2 Duet AI for developers Avenging Polayni's Revenge X/TWITTER: @DynamicWebPaige (Paige) @labenz (Nathan) @eriktorenberg @CogRev_Podcast SPONSORS: NetSuite | Omneky NetSuite has 25 years of providing financial software for all your business needs. More than 36,000 businesses have already upgraded to NetSuite by Oracle, gaining visibility and control over their financials, inventory, HR, eCommerce, and more. If you're looking for an ERP platform ✅ head to NetSuite: http://netsuite.com/cognitive and download your own customized KPI checklist. Omneky is an omnichannel creative generation platform that lets you launch hundreds of thousands of ad iterations that actually work customized across all platforms, with a click of a button. Omneky combines generative AI and real-time advertising data. Mention "Cog Rev" for 10% off. Music Credit: GoogleLM

Today, we're excited to get to know Vivek, AI researcher at Google and one of the lead researchers for Med-PaLM2, and Viswesh, CTO and Founder of Valar Labs! Vivek is a Research Scientist at Google Health AI advancing biomedical AI to help scale world class healthcare to everyone. Vivek is particularly interested in building large language models and multimodal foundation models for biomedical applications and leads the Google Brain moonshot behind Med-PaLM, Google's flagship medical large language model. Med-PaLM has been featured in publications such as The Scientific American and Forbes. Vivek graduated with his masters from UT Austin in Computer Science and Bachelors at National Institute of Technology in India. Viswesh is the CTO and Co-Founder of Valar Labs. Valar Labs is building clinical grade deep learning to analyze each patient's characteristics and provide clarity to oncologists during decision making. Their AI is built with oncologists at the center and provides interpretable and actionable insights. Prior to founding Valar Labs, Viswesh was a Research Assistant in Stanford's Artificial Intelligence Laboratory (SAIL) leveraging cutting edge artificial intelligence to solve healthcare problems. He was also the founder of Kanna, a patented and clinically-validated method to detect Amblyopia in children in India. Viswesh graduated with a bachelors in Computer Science at Stanford. In this episode, Vivek and Viswesh shares how they got into Healthcare AI research and how they fell into different career paths, one leading Research at Google Health AI and the other as the CTO and Co-Founder of Valar Labs. We talk about the future of LLMs in healthcare, and also how to build defensibility in AI healthcare startups.

Today, I'm talking to Demis Hassabis, the CEO of Google DeepMind, the newly created division of Google responsible for AI efforts across the company. Google DeepMind is the result of an internal merger: Google acquired Demis' DeepMind startup in 2014 and ran it as a separate company inside its parent company, Alphabet, while Google itself had an AI team called Google Brain.  Google has been showing off AI demos for years now, but with the explosion of ChatGPT and a renewed threat from Microsoft in search, Google and Alphabet CEO Sundar Pichai made the decision to bring DeepMind into Google itself earlier this year to create… Google DeepMind. What's interesting is that Google Brain and DeepMind were not necessarily compatible or even focused on the same things: DeepMind was famous for applying AI to things like games and protein-folding simulations. The AI that beat world champions at Go, the ancient board game? That was DeepMind's AlphaGo. Meanwhile, Google Brain was more focused on what's come to be the familiar generative AI toolset: large language models for chatbots, and editing features in Google Photos. This was a culture clash and a big structure decision with the goal of being more competitive and faster to market with AI products. And the competition isn't just OpenAI and Microsoft — you might have seen a memo from a Google engineer floating around the web recently claiming that Google has no competitive moat in AI because open-source models running on commodity hardware are rapidly evolving and catching up to the tools run by the giants. Demis confirmed that the memo was real but said it was part of Google's debate culture, and he disagreed with it because he has other ideas about where Google's competitive edge might come into play. We also talked about AI risk and artificial general intelligence. Demis is not shy that his goal is building an AGI, and we talked through what risks and regulations should be in place and on what timeline. Demis recently signed onto a 22-word statement about AI risk with OpenAI's Sam Altman and others that simply reads, “Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.” That's pretty chill, but is that the real risk right now? Or is it just a distraction from other more tangible problems like AI replacing labor in various creative industries? We also talked about the new kinds of labor AI is creating — armies of low-paid taskers classifying data in countries like Kenya and India in order to train AI systems. I wanted to know if Demis thought these jobs were here to stay or just a temporary side effect of the AI boom. This one really hits all the Decoder high points: there's the big idea of AI, a lot of problems that come with it, an infinite array of complicated decisions to be made, and of course, a gigantic org chart decision in the middle of it all. Demis and I got pretty in the weeds, and I still don't think we covered it all, so we'll have to have him back soon. Links: Inside the AI Factory Inside Google's AI culture clash - The Verge A leaked Google memo raises the alarm about open-source A.I. | Fortune The End of Search As You Know It Google's Sundar Pichai talks Search, AI, and dancing with Microsoft - The Verge DeepMind reportedly lost a yearslong bid to win more independence from Google - The Verge Transcript: https://www.theverge.com/e/23542786 Credits: Decoder is a production of The Verge, and part of the Vox Media Podcast Network. Today's episode was produced by Jackie McDermott and Raghu Manavalan, and it was edited by Callie Wright. The Decoder music is by Breakmaster Cylinder. Our Editorial Director is Brooke Minters, and our Executive Producer is Eleanor Donovan.  Learn more about your ad choices. Visit podcastchoices.com/adchoices

Artificial intelligence (AI) is dominating the headlines, but it's not a new topic here on Exponential View. This week and next, Azeem Azhar shares his favorite conversations with AI pioneers. Their work and insights are more relevant than ever. DeepMind's co-founder and CEO, Demis Hassabis, joined Azeem in 2020 to explore his company's progression from gaming to accelerating scientific discovery. In 2023, DeepMind's parent company Alphabet announced consolidation of its biggest research units, DeepMind and Google Brain, into a new division led by Demis.

Today we're joined by Nicholas Carlini, a research scientist at Google Brain. Nicholas works at the intersection of machine learning and computer security, and his recent paper “Extracting Training Data from LLMs” has generated quite a buzz within the ML community. In our conversation, we discuss the current state of adversarial machine learning research, the dynamic of dealing with privacy issues in black box vs accessible models, what privacy attacks in vision models like diffusion models look like, and the scale of “memorization” within these models. We also explore Nicholas' work on data poisoning, which looks to understand what happens if a bad actor can take control of a small fraction of the data that an ML model is trained on. The complete show notes for this episode can be found at twimlai.com/go/618.