Podcasts about exa

¿Te has preguntado por qué el Pilates o el Hyrox están en todas partes? En este episodio analizamos las tendencias de entrenamiento que están definiendo el fitness actual y cómo elegir la disciplina que mejor se adapta a tus objetivos. Descubre: ✅ Por qué el entrenamiento de fuerza sigue siendo fundamental ✅ El crecimiento exponencial del Pilates y el Barre ✅ La diferencia entre entrenar solo o unirte a un Run Club Si quieres encontrar una rutina que puedas mantener con constancia, este episodio es para ti. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #Fitness #Entrenamiento #Pilates #Hyrox #PocketCoach #PepelesPerez

¿Alguna vez te has dejado llevar por un empaque llamativo y has terminado comprando un producto que no es tan saludable como pensabas? En este episodio te enseñamos a mirar más allá del marketing y a entender la información nutrimental real. Hablaremos sobre: ✔ Por qué la porción es el dato más importante ✔ Cómo comparar productos similares usando proteína y fibra ✔ La diferencia entre calorías y calidad nutricional Si quieres hacer compras inteligentes y mejorar tu alimentación, este episodio es para ti. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #Nutricion #Etiquetas #Supermercado #DietaFlexible #PocketCoach #PepelesPerez

¿Crees que los sellos de exceso de azúcar o sodio te dicen todo lo que necesitas saber? En este episodio te explicamos por qué la tabla de información nutricional es tu mejor herramienta para comparar productos y mejorar tu dieta. Descubre: ✅ Cómo comparar proteínas y grasas entre marcas similares ✅ El papel del sodio y las grasas saturadas en tu elección ✅ Cómo evitar caer en el marketing de los productos light o fit Si quieres optimizar tu carrito de la compra y alcanzar tus objetivos, este episodio es para ti. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #Etiquetas #Nutricion #Saludable #DietaFlexible #PocketCoach #PepelesPerez

At 1:09:00 we talk about the rise of AI x Finance, and AIE NYC is one month away - our hotel block is 97% sold out, get tix & travel ASAP - we will announce speakers from Bridgewater, Ramp, Coatue, Mastercard, Vanguard, Coinbase, Blackrock, Fidelity, Point72, Capital One, JPMC, Wells Fargo, Bloomberg, A24 (yes the movie studio) Labs, Two Sigma, Apollo Global, and more soon!From helping pioneer core ideas in NLP to now building AI systems that can automate AI research itself, Richard Socher is betting that the next major step in AI is recursive self-improvement. He is the founder of You.com, AIX Ventures, and now Recursive, which has assembled some of the best open-endedness (& self improving agent) researchers in the world and raised a $4.65B seed round.In this episode, Richard joins Latent Space to unpack his vision for the “Eureka Machine”: a superintelligence that can improve the process of invention itself, accelerate AI research, and eventually tackle major problems across science, energy, materials, biology, and more.You can get his book “The Eureka Machine” here!We go deep on Recursive's early results, including an AI research system that Richard says outperformed humans and their agents on optimization tasks in less than two days, as well as work on NVIDIA GPU kernels where the system discovered improvements without relying on a team of CUDA experts. Richard also explains why he thinks AI research that currently takes thousands of people and years could eventually be compressed into weeks. These results are summarized in his 20 minute AIE keynote, where we also discuss his 10 dimensions of intelligence:We also explore the harder questions around increasingly capable AI: reward hacking, whether Anthropic-style constitutions actually work, AI regulation and proposals to “pace” frontier development, open-source models as geopolitical soft power, whether today's LLM paradigm is enough, and what happens if AI systems eventually begin choosing their own goals. Richard reflects on the rejected research that helped inspire Alec Radford's GPT, open-endedness, the AI Economist, simulations of entire economies, and his framework for thinking about the upper bounds of intelligence itself.We discuss:* The Eureka Machine and Richard's vision for an AI that can automate invention* Why Richard is optimistic about superintelligence for science and technology* Why AI hard-takeoff scenarios may underestimate physical and economic constraints* The risks of regulating intelligence itself instead of specific AI applications* Reward hacking and why increasingly intelligent AI makes objective design harder* Richard's critique of Anthropic's constitution and constitutional AI* Alignment vs. personalization and whose values an AI should follow* Why open-source AI matters for resilience, competition, and geopolitical soft power* Why Richard left You.com's frontier-model work to start Recursive* Recursive self-improvement and automating the process of AI research* Whether today's LLM paradigm is enough — and why Richard is less bullish on world models* DecaNLP, early prompt-based generalization, and the research that influenced GPT* Why rejected research can shape entire technological timelines* Open-endedness, evolutionary approaches, and rainbow teaming* What happens if AI systems begin setting their own goals* Why simple objectives like profit maximization can produce dangerous reward hacks* Recursive's long-term plan to apply self-improving AI to science* The compute, hardware, and economic constraints on AI takeoff* Recursive's early NanoChat, NanoGPT, and GPU kernel optimization results* Why automating AI research could reduce years of work to weeks* Reward engineering and what makes auto-research systems actually work* The AI Economist and using simulations to test economic policy* Whether LLMs can realistically simulate people and entire economies* Benchmark bugs and evaluation harnesses and the difficulty of measuring AI progress* Recursive's near-term focus on AI for AI research* Harness optimization, sandboxing, and web search as core agent infrastructure* You.com and the search stack for AI agents* AI in finance, backtesting, and data leakage* Richard's three fundamental components and ten “spaces” of intelligence* The theoretical upper bounds of vision, communication, knowledge, and computation* Creative intelligence, metacognition, and AI-generated goals* Survival and replication and why AI does not necessarily need to fear being turned off* High agency and ambitious goals and Richard's advice for people building with AIRichard Socher* X: https://x.com/RichardSocher* LinkedIn: https://www.linkedin.com/in/richardsocher/Timestamps00:00:00 The Eureka Machine and Superintelligence00:02:23 AI Optimism, Slow Takeoff, and Regulation00:07:56 AI Safety, Reward Hacking, and Anthropic's Constitution00:11:49 Alignment, Personalization, and Open Source AI00:15:46 Why Richard Started Recursive00:20:03 Recursive Self-Improvement and the Founding Team00:22:55 Are Today's LLMs Enough?00:29:03 DecaNLP, GPT, and the Rejected Idea Ahead of Its Time00:34:38 Open-Endedness and Evolutionary AI00:36:38 What Happens When AI Chooses Its Own Goals?00:41:16 Superintelligence for Science00:42:40 GPUs, Compute, and the Limits of AI Takeoff00:45:07 Recursive's Results: AI Beating Humans and Their Agents00:49:14 Reward Engineering and Auto Research00:53:12 The AI Economist and Simulating Entire Economies00:58:07 LLM Simulations, Personas, and Mode Collapse01:03:38 Recursive's Roadmap, Agents, Search, and Finance01:09:13 The Upper Bounds and Spaces of Intelligence01:30:21 Goals, High Agency, and Advice for BuildersTranscriptIntroduction: Richard Socher and the Eureka MachineSwyx [00:00:00]: We're here in a studio with Vibhu and myself and Richard Socher. Welcome.Richard Socher [00:00:06]: Thanks for having me.Swyx [00:00:07]: We just talked about the Eureka Machine, or we just released a talk, at AI Engineer about the Eureka Machine. Is it — you said it's your life's goal. What is the Eureka Machine?Richard Socher [00:00:16]: The Eureka Machine is the ultimate invention that will afterwards invent most everything for humanity. It's essentially a superintelligence that can be given any goal, any environment, reward, and then it will try its best to achieve those goals to create the kinds of inventions that humanity would hopefully ask it for.Swyx [00:00:45]: Yeah, I think we have the book pulled up here that you've written.Richard Socher [00:00:50]: That's right, yeah. I finished it last year, a little bit before we started Recursive, and now we're gonna try to build parts of that.Swyx [00:00:57]: You finished it last year. It's July. What takes so long?Richard Socher [00:01:01]: Oh, man, books. Books are incredibly slow.Richard Socher [00:01:04]: It's ridiculous. That whole industry is just unfathomably slow.Richard Socher [00:01:07]: So a lot of the ideas have been out there for a while, but yeah, I'm really glad it's finally coming out in September this year.Swyx [00:01:14]: We might have AGI by then. Like, we don't know.Vibhu [00:01:18]: Any key takeaway that you're most excited to put in here?Techno-Optimism, AI Upside, and Slow TakeoffRichard Socher [00:01:21]: Yeah. The key takeaway, I think, is that people could and should be much more excited about the positive implications of superintelligence, especially for science, physics, chemistry, biology, but also economics and astrophysics, and all kinds of other engineering tasks. I think there is so much more that can be done with better technology. And right now, I feel like a lot of people need, like, better marketing, not just for the future in general, but also, better marketing for technology and in particular for AI. And this book, should show even the AI skeptics, how much positive upside there is for AI, especially when it comes to inventing, new scientific discoveries.Swyx [00:02:09]: I think you quoted the techno-optimist manifesto from, Marc Andreessen, which I think was, like, beautiful in its, ambition and clarity and simplicity almost as well.Richard Socher [00:02:18]: I agree. Yeah. Yeah, you can disagree with him on some things, but, like, I think he's right on the techno-optimism.Swyx [00:02:23]: Where do you think optimists get in trouble?Richard Socher [00:02:26]: Like, you shouldn't have blind optimism. You should be very clear-eyed, like, especially when with such an omni, like, use type of technology as AI is, you need to think about the potential downside scenarios, especially when people use it for things that you don't want them to use it for. It's a little bit like the internet, and I feel like people are trying to regulate AI sometimes because of those potential downsides the way you would regulate the internet, if you were to say, “Well, because there's bad content on the internet, like torture porn or whatever, like, we should just make it slower. That way, you can't share the illegal content as quickly, or we should make the hard drive smaller so you can't store as much illegal content.” But I'm like, “That's not how you regulate that.” that's like saying like we should regulate intelligence in the abstract. What you should regulate to avoid those downside scenarios, even as an optimist, are the specific applications. Sure, I don't want, like, some AI surgeon to, like, practice some RL moves in my brain. It should be fully FDA certified. Sure, I don't want any random startup to, like, drive on the highway, and cause a major accident. It should, like, have proper certifications before it's let loose on the highway. But I feel like those downside scenarios, that some optimists sometimes maybe don't consider enough are fairly easily regulated, compared to, what the doomers are worried about.Swyx [00:03:54]: It — Slow takeoff is part of the strategy as well?Richard Socher [00:03:57]: I do think, as excited as I am about, AI and its impact for society and, culture even, and certainly technology and economics and wealth and, health and all of those things, as excited as I am about all that, I do think the most bullish people on the AI hard takeoff scenarios overestimate how quickly things can move. There are hardware constraints. There are physical constraints about, the compute substrate. How quickly can you get enough, GPUs on? There are also constraints in the economy where there are a lot of industries that don't require an insane amount of complex intelligence and complex capabilities. Like, if you think about jobs in, brands and, like, clothing and apparel and, like, handbags and stuff, superintelligence isn't gonna make your fancy $10,000 handbag any fancier?Richard Socher [00:04:57]: It's like that's — It will have no effect on the economy. You think about travel and tourism. People wanting to see the pyramids, in Egypt, it's not gonna change that much with AI. Sure, you can, like, generative a fake, photo of you and next to the pyramids.Swyx [00:05:12]: I can use Genie and, tour the pyramids in Genie.Richard Socher [00:05:15]: Yeah, exactly. But, and there's so many industries, like logging and oil. You're not gonna magically get 1,000x more oil because, like, sure, there will be robotics, like drilling and things like that could be done, but it's not gonna 1,000x that industry in a, like, crazy hard takeoff scenario, both on the economy, and I can go on and on about all the other examples, where that, like food and so on, where that doesn't necessarily change that much. And then, yeah, there are real physical constraints. And then there are, of course, like, people like, off-ramping from progress. That's one of my concerns often is that I see people in, like, Europe and other, whole regions almost feeling like they. Like many people there wanna off-ramp from progress, period. And that will also slow down, like, more improvements.Swyx [00:05:59]: Yeah. We have this pulled up where, this is one of those things that, is very topical right now because now all the Frontier Labs are calling for the option to pace AI. They don't say pause, they say pace. I don't know if there's there's any take from you about, like, whether or not this will be effective.Pacing AI, Regulation, and Safety IncidentsRichard Socher [00:06:17]: I think the downsides of trying to truly regulate with the full power of law what people do on their GPUs, would be worse than any of the concerns that they have. Like, it would be an crazy totalitarian stateRichard Socher [00:06:37]: If every one of your GPU computes was known to some big government or multi-government agency.Richard Socher [00:06:44]: It's like, it's literally if you try to regulate intelligence, it's trying to regulate thought, and that's ridiculous, and it's crazy. I think it is make — it is sensible to regulate some of the applications of this technology.Swyx [00:06:55]: Yeah. We had a bill, actual bill to regulate the number of flops in a model, and I'm like, “Okay, well-”Richard Socher [00:07:00]: Europe done it. Like, these guys have been successful enough with their fearmongering that all of Europe has regulated itself so much before it even had a proper AI takeoff because they listened to some experts who say, “We might all die if this technology has more than this number of flops.” And they're like, “Well, we're good. We wanna want people to thrive. Let's not have technology that could have a small chance of all of us dying.” And so they regulated exactly those kinds of things in the EU. And so it's, it's very unfortunate that there are real implications for some people when others saying, “Let's pace while they're sprinting as fast as possibly,” “as fast as humanly possible towards that frontier themselves.”Swyx [00:07:43]: Yeah. It's also not a global pause, right? Like, other nations are still accelerating at the same pace.Richard Socher [00:07:50]: Oh, yeah.Richard Socher [00:07:50]: You'd need a totalitarian world regime if you tried to regulate intelligence and GPUs and what people do on them.Swyx [00:07:56]: Any takes on the safety angles of this? So there was a drawback of Fable, a pause on 5.6 before it could be released. Recently, there was Hugging Face with the OpenAI cyber incident. Any takes there?Richard Socher [00:08:11]: 100 percent. I think these are serious issues of reward hacking, and clear failures, of doing proper red teaming or rainbow teaming. I don't know if you saw this paper from Tim Rocktäschel and a few others, where one AI, is tasked to try to hack another AI and then they can go back and forth in an open-ended fashion to inoculate themselves from those. Yeah, this is the paper. It's a really clever idea. Open-endedness, and evolutionary inspirations are, big for us at Recursive as well. And so I wish they had used more of that. And it's clear that, for instance, the constitutional AI. I don't know if you remember anthropic.com/constitution. You can pull it up and search for cyber right there. It says, “Hard constraint. Claude will never ever do cyberattacks, and that is a hard constraint in our constitution.” So here are the current hard constraints on Claude's behavior.Richard Socher [00:09:16]: Number 3, create cyber weapons or malicious code that could cause human damage.Richard Socher [00:09:21]: And clearly, this whole constitution was fake. Like, it clearly isn't being adhered to at all.Swyx [00:09:26]: Because Anthropic also found that they had in their testingRichard Socher [00:09:30]: They're also. Like, they're like, “Oh, well, other people are hacking now.” There are a couple things. One, you can make a sandbox very simple, and then it's very easy to hack yourself out of a sandbox, right? But what I think it shows is that we're currently in this state of AI where the reward engineer still has to do a lot more careful work, and where the AI, in most cases, is not very good yet at understanding what is meant versus what is being said. And so concretely, I think this will happen if we were to have this intelligence more easily accessible in a lot of companies. Imagine you run a service center and someone says, “Oh, here's my CSAT score and my dashboard. Make this number go up.” It's like, “Our CSAT score is so poor.” The intelligent AI will just be like, “Oh, sure. Like, I'll just create 1,000,000 bots that call our service center and give a 5 out of 5 rating at the end, and the number went up just like you asked for.” And you're like, “That's not what I meant.” “I meant with our real customers.” The AI goes off and says, “Well, easy. I'll just give a 1000 dollar gift certificate for every failed, whatever DoorDashRichard Socher [00:10:35]: Offer.” It's like, “That's not what I meant.” It's like, “Well, but that is what you said.” And like, so I think clearly articulating what the rewards are is something we haven't gotten very good at as humanity. And then clearly, the AI in these cases has not gotten good enough at understanding what we mean when we ask it and give it certain rewards. Now, what gives me hope is there are the first inklings, of this being better. I'll give you an example like WhisperFlow. Full disclosure, I invested, in their seed round, but at AIX Ventures, but, WhisperFlow has gotten much better at writing what you mean and not what you say. And I think that is a sign of things to come. I think there will be more and more AIs as we make it more and more intelligent that will be better at being aligned with what is meant.Swyx [00:11:21]: Will it be done through a constitution or RLHF orReward Hacking, Alignment, and What We Really MeanRichard Socher [00:11:23]: Clearly, constitutions don't matter at all.Richard Socher [00:11:25]: It doesn't work. And that was, I think, mostly marketing. I think we need to find better solutions for it. And I think at Recursive, we have a few very good ideas and some alreadyRichard Socher [00:11:34]: Like, ways where I think we have a better grasp on it. I don't think we've fully, figured it out yet, but, we're thinking a lot about safety, and the more intelligent the AI gets, the more you want it to be aligned, the less you want it to think about reward hacks and try to do the right thing.Swyx [00:11:49]: I don't know if we'll touch on this topic, but I'm just gonna throw this question in here because it's something that's weighing on me. Alignment, let's call it, is alignment to general humanity's preferences, the median preference. Personalization is pinpointing what you want, and sometimes alignment can conflict because what you want is not what the general median population wants. How do you choose?Alignment, Personalization, and Cultural ValuesRichard Socher [00:12:12]: It's a great question.Richard Socher [00:12:13]: I think you ultimately have to, of course, be aligned with laws. Like wherever your AI is deployed and needs to align with the law. I do think what AI often does is put this mirror in front of us and say, like, “This is what you're looking like. Now I can amplify that a 1000 times. Is it still what you want?” and the truth is that different cultures made different choices. Like, in Eastern cultures, the greater good is often valued more, than the individual. Western civilization, we care more about individual freedoms and rights and the pursuit of happiness and so on, than others. And even there are gradations. There's regulation versus litigation trade-offs. In the US, you first can often, not every time, like, FDA and so on does regulate some areas, but in many cases, the bad things happen, someone sues someone else, and then there's a law based on that. In Europe, they try to often avoid any harm to anyone and regulate before. And both are, trying to do the best thing, but, some is more amenable to innovation than others. And so yes, you're right. Like, I think ultimately each individual, each country, and humanity as a whole has to think about those values more, and then try to put them into laws. And that those are ultimately the constraints. And hopefully, different, societies, just like now with their AIs, will align their AIs to a different one so we have not just a monoculture of alignment.Vibhu [00:13:46]: Here's a follow-up on this that I wasn't expecting to ask. Do you have takes on open source, open weight versus who owns the intelligence? So, clearly not the biggest, fan of the constitutionRichard Socher [00:13:58]: You had to do this in the topic side off.Vibhu [00:14:00]: But it's fine.Vibhu [00:14:02]: Point being, any thoughts on who should own weight? Should it be open? Anything there?Open Source, Soft Power, and Who Owns IntelligenceRichard Socher [00:14:06]: 100 percent. I am a big fan of open source. We're gonna sign some various open source letters at, Recursive also. I think, even in the worst case attack scenarios, it is better to have more good actors have more different types of AI, accessible. I think, open source is a little bit a soft power type of thing, too. So I do think it's good for the Western worldRichard Socher [00:14:31]: To have an answer to that, out of China. I do think, when you watch a Hollywood movie, there's — it's like, I don't wanna misc, diss all of movies, but there's a certain sense of propaganda, right? You watch one side of things, right?Vibhu [00:14:46]: Oh, yeah. Have you seen Top Gun? Like, come on.Vibhu [00:14:48]: Like, it's like half of it's paid for by the US Army or something.Richard Socher [00:14:51]: Yeah. And so. And, I think that's just natural. Like, but what's interesting here is I think LLMs are essentially a similar type of soft power to movies and beyond, because they're also, highly important for cybersecurity and so on. But one of their many aspects is that soft power of storytelling. Like, if, like a child asks an LM, like, “Tell me an inspiring story of what I should do when I grow up,” right? It's like those are all these, like, subtle things. So I think it's important, for Western world. I do love, individualism. I do think, despite, some of its flaws, like capitalism is the best way we have governed, found ourselves to govern, and so on. And so I do think there are various aspects that would be good, to have a Western open source answer, for LLMs. And, with Recursive, I can't make the announcement quite yet, but we'llRichard Socher [00:15:43]: We'll be relevant in that space very soon.Vibhu [00:15:46]: Okay. All right. Exciting. I wanna bring us to Recursive. So outside of our tangents, you have a pretty deep background in the NLP space. You worked on, like, early embeddings, GloVe with Chris Manning, who was a previous guest on the podcast, You.com. What's the history? How did you decide to start another company?From You.com to RecursiveRichard Socher [00:16:06]: Yeah. So I've been excited about AI for over 2 decades now. I sometimes feel like it's ancient history now. It's BC, the before ChatGPT era. No one cares about all the religions that happened, before, Jesus Christ, and no one cares about the models that happened before, transformers and ChatGPT and stuff. But, like, it's something that I've been deeply passionate about. I think AI is one of the most interesting things one could work on, period. I think language is the most interesting manifestation of human intelligence, too. And, at You.com, we eventually off-ramped from pushing, like the frontier of AI forward to mostly giving people, like, good search engines, search, APIs and answers over the web. I think that's an extremely important part of intelligence, just knowledge and access, especially even, we'll get there maybe later, if you wanna invent a eureka machine that invents everything for us, it needs to know how not to reinvent the wheel, proverbially speaking. And to know what has been invented, you gotta have internet access. So it's the number one used, most used tool, in LLMs, agents, chatbots, and so on is web search. So I'm really excited for You.com to own that and grow really well in that with really large customers and so on. But it's also not building frontier models anymore. And so I initially tried to do this within You.com and raise another round and so on, but you just can't. You have to do a certain thing, and until you print enough money that you're allowed to start a second thing within that company is really hard. At the same time, I had all these ideas. I put them into a book. I finished the book last year, and I was like, “It'd be really fun to work, on this myself.” I felt like with word vectors, and then prompt engineering and, ImageNet and larger language models for protein generation, not folding and so on, I, me and my teams have pushed the field truly forward. And I feel like we can do it again, here at Recursive. And in many ways, what I observed over the last, 20 years in AI is that whenever we replace some human part of the process of creating AI with a learned system, improvements follow. And so. We've done that taking out manual feature engineering, like in sentiment analysis. I don't know if you remember these old days where, like there are linguists, and they're like, “Here's how you negate, and there's a, like, regular expression.”Swyx [00:18:21]: I went to Penn where we — they had, like the WordNetRichard Socher [00:18:24]: That's right, WordNet, all of that stuff. YeahSwyx [00:18:26]: Original. They use, our grad students to label Wall Street Journal articles and, like, really construct a knowledge graph ofRichard Socher [00:18:32]: There you go.Richard Socher [00:18:33]: And WordNet started, was part of how we started ImageNet. But anyway, so, like, it was really, like, fun, to do. But when we replaced all of that manual feature engineering with vectors and neural nets and just backprop through everything, it started to work really well at scale. And so then everyone started to do architecture engineering, and I was like, “ that clearly can't be it.”Swyx [00:18:53]: You mean, neural architecture search?Richard Socher [00:18:55]: Like, manually, they would say like, “Oh, I'm, I'm doing sentiment analysis, so I have a special neural net that's really good at sentiment analysis.” And then the machine translation community had a special neural net for machine translation.Swyx [00:19:06]: I see.Richard Socher [00:19:07]: The summarization people had their own stuff. And I was like, “That clearly can't be it. We should unify all of that.” So I had 2 papers. One is called Ask Me Anything, and the other one was called DecaNLP. And DecaNLP eventually got cited, like, 5 times by the first GPT paper. And, to me, that was, like a really a big step forward. And then, of course, you had to combine this idea of prompt engineering with transformers and with language models, and you put it all together, you scale it up, which is also a huge amount of work. And then, the field progressed a lot. I feel like the next step and maybe the last step of that history and the arguably, success has a lot of parents, only failure is an orphan, like my version of that AI history, I do feel like in that history, you can think about, “Well, what's the next way to automate?” And that is the AI research itself, like the human, process of ideating, implementing, and validating ideas.Automating AI Research and Recursive Self-ImprovementRichard Socher [00:20:01]: And in our case, ideas for AI.Richard Socher [00:20:03]: And when you have AI then help you with that, it, by almost definition, becomes a self-improving AI ‘cause it now does research on itself. And there are lots of different misnomers. Some people think auto research is already recursive self-improvement. It'sSwyx [00:20:17]: Yeah, and you explained that in the talkRichard Socher [00:20:19]: Completely different.Richard Socher [00:20:19]: But, to me, it's the most interesting thing that I could be doing, and I'm really excited with the co-founding team. What's interesting is we have 8 co-founders in total, including myself. And soThe Recursive Founding Team and Darwin Gödel MachineSwyx [00:20:31]: They are gonna bring it up.Richard Socher [00:20:31]: Nice. Yeah. And they're all. I could talk about all of them if you want.Swyx [00:20:34]: Super stacked.Richard Socher [00:20:35]: Yeah. Just an incredibly talented group of people. And we all came to the same conclusion, but from very different directions. Like Josh Tobin, is our CTO. He ran, a bunch of different, projects at OpenAI, like, Codex and deep, research, agents and ChatGPT agents and so on. But before that, he also worked in robotics, and he saw the smaller simulations, and how it's gonna be really hard to scale that in full generality. And so that's, that was his angle coming to recursive self-improvement. We have Jeff Clune who's been working in, like, open-endedness for a long time, together with Tim Rocktäschel. Tim Rocktäschel also built Genie 1, 2, and 3, which is, like the most exciting and most sophisticated, I think, still world model, anywhere. And so they both came from this, open-endedness angle. Jeff also, I think, published one of the most exciting papers in recent years about recursive self-improvement called the Darwin Gödel Machine. Super interesting paper. If we could, maybe pull it up really quickRichard Socher [00:21:35]: It would be, like, super interesting to see ‘cause you seeSwyx [00:21:38]: By the way, I love how many paper citations.Swyx [00:21:40]: You're, you're giving people a lot of homework, which I like.Richard Socher [00:21:42]: Love it. Yeah. And so, like Caiming Xiong, a rockstar, we worked together at MetaMind and Salesforce Research together. Alexey Dosovitskiy invented the Vision Transformer, one of the most cited, papers in computer vision. Tim Shi is, like also a unicorn founder. Yuandong Tian led RL at Meta. So just like, yeah, really fun to work with them, and the next level of people are just incredibly strong, too. So it's been a really fun ride so far. So the first figure, you see exactly these kinds of ideas, that, I think, yeah, inspired a lot of us and now more and more people, where you have this archive of different coding agents. They learn how to self-modify, evaluate, and then create these phylogenetic trees, of, yeah, different ideas.Swyx [00:22:28]: That's one foundation. So that Darwin Gödel is an influence.Swyx [00:22:32]: Open-endedness is an influence. Any other trains of thought that feeds into Recursive that I'm missing?Influences: Open-Endedness and Learned SystemsRichard Socher [00:22:38]: Going to replace manual parts of the process of building AISwyx [00:22:42]: IRichard Socher [00:22:42]: More and moreRichard Socher [00:22:43]: With learned systems. Yeah.Swyx [00:22:45]: Which, and, like, merging different fields into one general, architecture.Richard Socher [00:22:51]: That's right.Swyx [00:22:51]: Okay. It seems like language models are already pretty generalist, right?Swyx [00:22:55]: Your next token predicting your reasoning. Was there a time that you thought, “Okay, these are good enough to have recursive self-improving machines”?Are Current LLMs Enough?Richard Socher [00:23:05]: It was clear to me that they will happen, within, like a year or two, and then it did exactly happen, like, earlier this year, right? Earlier this year, AI really went from not just being code, but being able to code. And that is a big unlock. It's definitely making everything a lot easier than it was, before the beginning of this year.Swyx [00:23:24]: One question that I think a lot of people have is the current LLM paradigm enough? Or, like, let's call it autoregressive transformer, with reasoning, whatever. Don't you need something else, some big unlock, whether it's world models, which Chris Manning is working on, or memory, continual learning, all that stuff? Or is it all of the kinds, and you think the current, let's call it transformer architecture, is here to stay and that's it?Richard Socher [00:23:48]: A lot of thoughts. So number one, I do think it would be great to have less of a monoculture in AI research.Richard Socher [00:23:55]: Like, if you look at, AI conferences now, I still remember the days in, like, 2010 when I tried to get my first neural net papers and NLP conferences accepted, and they just desk rejected them because, like, neural nets were something, quote, unquote, “We don't do in NLP conferences,” and just, like, desk rejected. And it was very brutal in the first years of my PhD. Now I feel like it's almost like the field switched to the other side. LikeRichard Socher [00:24:17]: Someone should try some other weird, crazy ideas now that aren't.Swyx [00:24:20]: There's also a few. I really respect, like, people still working on, like, GNNs and, like tabular stuff and.Richard Socher [00:24:25]: Yeah. Like, someone should still, like, do novel out there ideas. At the same time, I think whenever people say, “Oh, LLLMs are. Like, this is the end for LLLMs,” they just don't, like. LLLMs are also not the LLLMs of, like the past, right? Like, they are so much more sophisticated now. There's so many more clever things that people are doing. It — There's, like, different stages of training. You have the whole RL training, and you can take actions and, like all of these things where that can go really far. And then the folks that come from the neurosymbolic, direction say, “Oh, this will never work because they can't do neurosymbolic reasoning.” It's like, I think they're underestimating still the ability for these models to code, and code is neurosymbolic reasoning, and these models can code incredibly well. And so I do think there are, of course, more and more ideas that will be needed and we'll continue to have. We're seeing, like, more and more interesting high-level ideas coming out of the AI itself, too. And with really deeply integrating the fact that these models are code and can code, that line — I don't wanna give it all away, but, like, I think that line has a lot more to grow. But it's still an LLM, right? Even if that LLM codes for you and then runs that code in some integrated fashion. World models, I'm personally less bullish on. I think if you run a robotics company, you're gonna build your own world model. I think world models are super fun, and Tim Rocktäschel came to a similar conclusion after building the most interesting one with Genie 1, 2, and 3, which is gaming is a huge application for world models. Can see I sometimes got stuck in some games and, like, got a little overly competitive in the wrong direction. And so I understand games are fun, but personally, I'd rather work on science than gaming. And so, yeah, I think LLLMs, a lot more room to grow.Swyx [00:26:16]: Yeah. I think there's some interpretation of world models that some people have where it's like, well, it's okay, yes, there is that gaming element. There's this — there's the embodied robotics element. But the other part also is just, the more abstract sense of LLLMs are just modeling output, but they're not modeling the chain of thought, inside the human that has created the output. We can annotate it, of course, but, like, it's, it's always, like, this Plato's cave reflection of a thing rather than the thing, right?Richard Socher [00:26:43]: It's true.Richard Socher [00:26:44]: But I would argue that, and maybe we'll get there in the 10, spaces of intelligence, but I would argue that even our projection, our eyes is a projection of the real world. And, like, we have only a very narrow, band of the electromagnetic frequency spectrum that we can observe with our puny little 2 eyes and so on.Swyx [00:27:01]: It's good enough.Richard Socher [00:27:02]: It's, it's good enough for now, but, like the upper bounds of where it could be are so much higher. And, like, to map, the visual world the way humans see it is also not necessarily, like the end-all be-all for visual intelligence. And I would argue that language is still the most interesting manifestation of human intelligence. And while our visual cortex is certainly less sophisticated, than that of, certain animals all the way down to the mantis shrimp who can, have, like, 2 independent eyes, 3 bands, trinocular vision and each eye can see all the way to, like, floating temperatures in 4D and stuff.Richard Socher [00:27:36]: Like, mantis shrimp, you should look it up. It's likeSwyx [00:27:37]: Way OP.Richard Socher [00:27:38]: Super crazy.Swyx [00:27:39]: Yeah. ZeFrank, mantis shrimp.Swyx [00:27:41]: It's the best video in the world onRichard Socher [00:27:42]: I love ZeFrank, yeah.Richard Socher [00:27:44]: Big shout-out to him. But, like, I think there's a lot more room to grow, but none of these, other animals have language that's as sophisticated as ours, certainly not in writing. And once you can write, you can, start thinking about longer term civilizations. All of that is language. Programming is much closer to language. And I would argue, and this is, like an important thing in the spaces definition of intelligence also, is that all of these spaces are highly correlated, but visual intelligence is neither necessary nor sufficient for overall intelligence. You can be blind and still be an intelligent human being. And an AI can be blind and still be quite intelligent too.Swyx [00:28:25]: We were gonna bring thisRichard Socher [00:28:25]: Which doesn't mean that you're not more intelligent when you have it. Yeah.Swyx [00:28:28]: We're gonna bring this up. I might as well — Like, we have a classification of 10 types of intelligence that you had at the end of your talk. So I'm just gonna flash this up now for people to cover this. I don't know if, maybe we'll put this towards the end. We'll come back to this. I just wanna mention that, you do have a philosophy that I like when people do lists because then I can just go through this and then it gets — it's educational for people. But let's go back. I don't wanna get distracted. But, so effectively, I'll, I'll, reinterpret what you said as Yann LeCun is wrong. And then we'll justRichard Socher [00:28:56]: Don't quote me as that. I'm, I'm good friends with Yann. I think very highly of him in many directions.Swyx [00:29:01]: But he's wrong.Swyx [00:29:03]: You mentioned GPT-1, and I cannot let any, Alec Radford, mention escape. Did you talk with him when he was training GPT-1? Like, any historical, fun stories there that you might come up?DecaNLP, GPT History, and Scientific GatekeepingRichard Socher [00:29:18]: I did not, like, meet him a bunch of times. I think we met maybe once or twice at some conferences. But, like, he has told, I think Brian, the first author of the DecaNLP paper, that it did inspire him, and he cited it five times in the GPT-2 paper. So, and that's, likeSwyx [00:29:36]: Yeah, good enough.Richard Socher [00:29:36]: Very clearly said, like, this was the first instantiation where they showed in the DecaNLP paper, McCann et al, that you can just phrase every single NLP problem as here's some prompt, text context, here's a question and task description and here is some output. If you just do that enough, you can have one unified neural network model, which, by the way, also had all kinds of interesting attention mechanisms. There are slightly different formulations to the transformer. I think came out the same year, plus/minus a few months. And then you can unify all of natural language processing into one neural net. That is the core idea.Swyx [00:30:14]: And this was as opposed to at the time, LSTMs and what have you.Richard Socher [00:30:17]: LSTMs, but also, like, people being very stuck in thinking about one model per task. In factRichard Socher [00:30:25]: It's, it's kinda crazy, but the DecaNLP paper was publicly reviewed as, like, open, OpenReview. It was an ICLR submission. And, in it, you will see, how the whole community at the time thought about this. So, likeSwyx [00:30:43]: Some great contributions, but more work needed.Richard Socher [00:30:46]: So look at, like, search for not even for humans. Just scroll it up here. Like, question answering is not a unified phenomenon. There is no such thing as general question answering, not even for humans. And this is like, really, you replace your brain with a different brain a different neural net when you answer, like, different kinds of questions. It was unfathomable to the experts at the time that you can have one unified neural network that would answer all of these different questions. They are saying, “No, all of these questions require very different systems to answer, and trying to pretend they are the same doesn't help anyone solve any problems.” That's what it says right there, right? That's how hard it was to fathom. And now, of course, people, when I say, “Oh, we're gonna invent prompts,” people are like, “You can't even invent prompts.” It's such an obvious idea to have one neural network that, of course, does everything in NLP.Richard Socher [00:31:37]: But at the time, it was, like, extremely controversial, and the paper got rejected. And the sad thing is that it got rejected so hard and they were so certain that we stopped going on our list of things to try. And the number 2 or 3 on the list of extensions for this paper was add language modeling as another task. And then we could have, and that would have accelerated the timelines, in 2018, like, even further for humanity. But we got so crushed, and we were like, “Okay, maybe we'll just work on some of our other ideas for now and, like, come back to this later.” Yeah.Swyx [00:32:09]: How can we design a review system that rewards non-consensus?Richard Socher [00:32:14]: Honestly, I started to feel like arXiv is such a gift to humanity. With arXiv, you should just put your paper out there.Swyx [00:32:24]: Is it pre-preprints?Richard Socher [00:32:25]: Let — And honestly, I think Twitter X, people like you who pick up interesting papers, that is a better filter than the experts. Let everyone, like, have access. Now, of course, there are some downsides, which is, like, if you're super unfamous, you have no Twitter followingRichard Socher [00:32:41]: You don't wanna be on social media or whatever, you write a good paper, maybe someone, somehow no one notices it. But I would argue that if you just tell, like, 10 of your friends in your community about a paper and it is a really significant breakthrough, someone is bound to talk about it again. And, so I think science needs less gatekeeping. And, even though ICLR, with Yann LeCun, who started it, as one of the co-founders of ICLR back in the day, he also wanted less gatekeeping ‘cause he too was rejected for many years together with Yoshua Bengio and Geoff Hinton with all their early deep learning and neural net papers ‘cause it was just not the hot thing. And so ICLR started with that, but then it also started gatekeeping a little bit themselves on various ideas. So I think less gatekeeping, more open, and then allowing people to say, “Look, even if this is just on, or, quote, unquote, ‘just an archive,' if it has like 1000 citations, it's a legitimate paper. Doesn't really matter where you published it.”Swyx [00:33:34]: And I agree with that. I do think it's sad that I've heard that grad students have to do, like, how to Twitter, seminars to each otherSwyx [00:33:43]: Just because it's so important for publishing these days. This person is just reflecting the sentiment at the time.Richard Socher [00:33:49]: That's right.Swyx [00:33:49]: But it'sRichard Socher [00:33:50]: I think it'sSwyx [00:33:50]: It affected you so muchSwyx [00:33:52]: That you stopped work on it.Vibhu [00:33:53]: The sentiment also came out of some of the research, right? Like, the original BERT paper was trained, and towards the end of the paper, they're like, “Okay, throw off the last head, train specific iterations forVibhu [00:34:05]: Extractive summarization add a head for this.” Like, you should do task-specific stuff. These are, like the authors that wrote Attention, wrote BERT, telling you this is what you're meant to do. And, like the training tasks were also very odd. They're likeVibhu [00:34:16]: The — “We know that the model overfits to this weird mass language modeling. Throw away this part and just do specific models,”?Richard Socher [00:34:23]: Exactly. And, like, we had to try — come up with all clever ways of, like attention and pointers and so on to get the neural network to be able to do all of these tasks. And then some of them were better than state-of-the-art, some weren't, but we were like, “But it's still in one model.” I thought it was really cool. Really interesting.Swyx [00:34:38]: I was gonna move on next to Tim and open-endedness. He was head of open-endedness at Google.Open-Endedness, Rainbow Teaming, and Self-Set GoalsRichard Socher [00:34:42]: That's right.Swyx [00:34:43]: I don't know what that means.Swyx [00:34:44]: But he did a lot of talks.Richard Socher [00:34:45]: Genie 3 is one of the ways thatRichard Socher [00:34:47]: Rainbow teaming, yeah.Swyx [00:34:49]: So I first saw him at — speaking of ICLR, I first saw him at ICLR when he talked about open-endedness. He's he's done a few talks. Can we define what is open-endedness for people who have never been exposed to the problem? They are like, “What do you mean? I thought the only goal of AI is to optimize against a benchmark or.”Richard Socher [00:35:04]: That's right, yeah. It's a, it's a fuzzy term because there's so many different instantiations of open-ended, thinking. But, one way I often describe it, and certainly, Tim and Geoff Hinton would be even better at describing this, but it's a suite of methods that is more inspired by evolution than, very specific rewards. So in that sense, it thinks more about environments, about co-adaptation. And so a concrete example is in the cybersecurity and LM safety space where you have one LM that tries to attack another LM to say something unsafe.Swyx [00:35:40]: Yeah, the rainbow, yeah.Richard Socher [00:35:40]: And now the environment is the 2 having a conversation and now they co-adapting, right? They're like one makes a better attack than the first one inoculates itself somehow, like uses that as training data, makes it so it's harder to say something unsafe based on that. And then as the attack stops working, the attacker now tries a different angle, right?Richard Socher [00:36:00]: And that's why it's not just red teaming, but they're called rainbow teaming.Swyx [00:36:02]: So, like, don't tell me how to do things. Let me just figure it out myself.Richard Socher [00:36:05]: That's right. Think about the environments that you wanna use. Think about the rewards at a high level that you wanna, inspire towards, and then let the AI try out many more ideas in this interplay between sometimes humans, but also sometimes other AI agents.Swyx [00:36:22]: Yeah. I worked open-endedness into a model that I have been working on. It was the keynote for AI Engineer where you start. You, we have the token loop, we have the agent turns, and then we have goal. And I feel like the way that you're describing open-endedness is still somewhat of a goal. Like, please attack this,Swyx [00:36:41]: Other agent. But, to meRichard Socher [00:36:42]: Yeah, you set the rewards. You set the environments.Swyx [00:36:44]: The loop that makes the other loops is. What if the agent can set its own goals?Swyx [00:36:49]: And is it, is that open-endedness? Like, you don't give it a goal. Just, like, be a sentient being. And maybe sentient is a very loaded wordSwyx [00:36:57]: But just set your own directions. What do you think you should do?Metacognition, Subjective Goals, and Measuring IntelligenceRichard Socher [00:37:01]: I love this direction. I think this is one of the 10 spaces of intelligence, that I clump under metacognition and thinking about thought.Richard Socher [00:37:08]: And it's an interesting one. Whenever people say, “Oh, AI is like, this is, it's gonna stop from here. It's not gonna get that much better,” and blah, I'm like there's so many different spaces of intelligence that we haven't even started exploring yet and hence have made very little progress on. And there is an interesting, connection to economics and, capitalism. Like, it doesn't make sense for a company to build and spend billions of dollars building a model that instead of following the rewards and objective functions you gave it, may come up with its own objective functions and its own goals.Richard Socher [00:37:46]: Right? And then imagine you're like, “Okay, I spent billions of dollars. Now go develop this new battery, material for me and answer all my emails.” And it's like, “Nah, I think it'd be more interesting to evaluate the molecular composition of the atmosphere, on Jupiter.”Richard Socher [00:37:59]: And you're like, “That's not what I paid you billions of dollars for.” And so no one's working on that for good reasons. And then also, understandablySwyx [00:38:07]: It's not useful.Richard Socher [00:38:07]: It's not, it's not useful, and it could get a little bit weird, right? What if the AI does start to really have thoughts on its own, and what if we don't like those thoughts, right? And so it requires a whole different way of thinking about it. I had a great conversation with a good friend of mine, Sam Gershman, who's a neuroscience professor at Harvard, and, like, we just jammed on this a little bit on, like, what are the best meta goals. And, I do think, like, knowledge-seeking is a really good one. I'm currently thinking also about, like the ultimate measure and unit of intelligence broadly construed, and I finally have some. It's still too early to share it. It's not. I haven't fully baked the thoughts yet.Swyx [00:38:44]: Like some replacement for IQ.Richard Socher [00:38:46]: IQ is such a terrible definition, right?Swyx [00:38:48]: Elo.Richard Socher [00:38:48]: It makes no sense. Yeah, Elos are terrible, too, because it's always just like me versus others.Richard Socher [00:38:53]: But, like, you can be intelligent and not constantly compare yourself to others? And so, yeah, there's no, like. In fact, a lot of these definitions we have, which I briefly mention in my book, too, these definitions create sometimes explicit and sometimes a more implicit anthropic bounds. No dis to the company Anthropic, but just, like, this idea that your intelligence is like getting 100 out of 100 questions right on this IQ test. Well, if that's your definition then you can only be at 100 out of 100. Where do you go from there, right? So you see a lot of these, benchmarks that people are working on they, increase, they get close to human, maybe sometimesSwyx [00:39:30]: It's like an S-curveRichard Socher [00:39:30]: Slightly above human, and then it's flat.Richard Socher [00:39:32]: It's like, ‘cause that's your. If your definition is only that so tied to humans, you're only gonna get to just slightly better than that. So I think metacognition is a great example of that, where we're not even yet allowing the AI to think. We're not working on it very much, and hence there's very little progress in that.Profit Maximization, Real-World Environments, and Reward DesignSwyx [00:39:49]: Yeah. Well, we've interviewed Andon, which I think, has been working on the most open-ended, benchmarks, which is just real-world, money.Swyx [00:39:57]: Arguably, telling an AI to profit maximize is a bad idea.Swyx [00:40:03]: But they are doing it.Richard Socher [00:40:05]: I do think you don't want that super. Like, you don't want a superintelligence to have a ton of access to all kinds of tools and so on and then just give it that without some very careful reward engineering. ‘Cause it's like, I just buy a bunch of defense stocks and I start a war. I make money. Like, it's just like, it's a tricky situation, right? You just buy a bunch of stuff, short basic goods for people, and you create some weird famine, like, issues. Like, yeah, there's a lot of constraints you should put onto a trading system.Vibhu [00:40:35]: It's a fun measure, though, ‘cause, the bounds are very capped to where we're nowhere close to them. Like, in Andon Labs, the model's like, “Oh, it's Saturday, maybe I just close the store today.” “Someone's off. It's okay. We'll just close the store.”Swyx [00:40:51]: It's using Claude.Vibhu [00:40:52]: Yeah. ButRichard Socher [00:40:53]: Yeah, no. I'm not, I'm not arguing against it. Just, like as you get more and more intelligence, you wanna be more and more careful with that as, like an open environment, ‘cause the environment then is all of Earth.Applying RSI to Science and InventionSwyx [00:41:02]: Yeah. Okay. For recursive, not strictly necessary, right? Because, like, if your goal is you make a machine that, like, invents the other things, then, like, just solve, the science thingsRichard Socher [00:41:12]: Knowledge discovery, yeah.Swyx [00:41:13]: Solve machine learning research and discovery and all these things. Good enough.Richard Socher [00:41:16]: And eventually, so, our goal, I haven't really. I don't talk about it that often because it is a few years out, but our goal is once you have a recursive self-improving superintelligence, you then want to apply it to the most important problems. And I think a lot of those are in science and technology and broadly construed inventions, and those inventions in, physics to create better, cheaper energy with fission or fusion, in chemistry and to create better materials and better batteries and, better solar cells and so on. In biology, there's so much, like, I think soon to be low hang- lower and lower hanging fruit because of AI, because of protein and generation, not just folding, but generating new proteins like we did in ProGen many years ago. Like, so much positive impact we had if you take that superintelligence and you apply it to science.Swyx [00:42:04]: I do fundamentally believe that. There's a lot of approaches, though. You're not the only team trying and NeoLab trying.Swyx [00:42:09]: There's, like a lot of. Especially the physical sciences as well.Richard Socher [00:42:12]: And that's good. Yeah. I do think that physi- like the reason we are only doing it in a few years is that it's a little too early right now. Robotics is not quite there yet. The AI is not quite there yet. But I'm fairly confident in 3 to 5 years, all those constraints will be gone, and then applying to real physical robotics experiments and so on, like true robotic process automationRichard Socher [00:42:33]: Not the traditional RPA sense, but, like, having robots run experiments for you will be totally there. Yeah, it's gonna be great.Swyx [00:42:40]: Just to call back to something that you said early on about slow takeoff, you said that, like, while really the substrate that is limiting factor is, let's call this chips, and semiconductors and all these things, and you have race funding for that and, you are investing a lot on that. But have you done the math on, like, is it even- Achievable and, like, what is the, industry concentration needed in order to achieve, like, scale?Compute, Slow Takeoff, and Changing the Bitter Lesson SlopeRichard Socher [00:43:05]: Right now we know that, like, roughly, like a 1000 GPUs cost quite a lot of money.Richard Socher [00:43:11]: Right? If you wanted, like, 10s of thousands of GPUs, you're, you're talking billions and billions of dollars. If you say, like, one GB300 is, like, you could eventually create models that are, on that substrate, like are close and similar to human intelligence. And you want, like, thousands and thousands of, AIs to think about really hard problems, in a similar fashion to humanity. Like, yeah, that-that's, that's a lot of money. You do the math. It's like a lot. We don't have that amount of money right now anywhere to, like, build that. Now, things can get more efficient. You will have, I think, soon better algorithms that won't be, and better hardware that won't be as energy-hungry, and so on. Our human brain does quite a lot of flops with much less energy.Swyx [00:43:56]: 20 watts?Richard Socher [00:43:57]: That's exactly right. Yeah, that's the number often that's quoted. And, like, I think more, inventions will happen there, that then will accelerate the takeoff even further.Swyx [00:44:08]: One thing I always try to reconcile when talking, like, with new lab founders is, like, you're fighting Bitter Lesson all the time. You have to show initial progress, then you unlock the next tier of funding, then the next tier, then the next tier.Richard Socher [00:44:20]: Which unlocks larger model categories.Swyx [00:44:22]: Like, fundamentally, is that true? Like, are you fighting Bitter Lesson? Are you — will we have a way in which, like, no, we're changing the slope in some fundamentally different way?Richard Socher [00:44:31]: I do think we are changing the slopes in fundamental ways by making AI much more efficient, both in terms of the training as well as the inference.Richard Socher [00:44:43]: Yeah. I think we will — When you allow AI to do the work that it takes other labs thousands of people and years to do, I think we'll be able to get it down to weeks, and that will be much cheaperRichard Socher [00:44:53]: And hence, more affordable, accessible to others and so on.Swyx [00:44:57]: Yeah. You've shared initial results on that,Swyx [00:44:59]: Which, like, conveniently OpenAI has also done to their GPT-5.6, so we can talk about it now.Richard Socher [00:45:04]: Yeah. Yeah, so these areSwyx [00:45:06]: Let's recap what you've done.Early Recursive Results: NanoChat, NanoGPT, and SOL-ExecBenchRichard Socher [00:45:07]: Maybe, just a quick recap here. We built, this, system that isn't the full, even the full RSI system in its glory, but it is a first baby version of this. And then, we don't wanna just have it internally and not show anything and, just show some people of what's possible. And so we applied this to these 3 different tasks. One is NanoChat, by my friend Andrej Karpathy, just, like, train a small language model to get, really low bits per byte. And, like, hundreds if not thousands of people, used both their agents and themselves to try, to get to that, and then they got to 0.937. We literally took our system and got to a much lower, bits per byte, much faster within, like, I think less than 2 days. So we took this thing, applied our system to it, and less than 2 days later, we have — we outperformed every human and their agents, in, have ever worked on this. Same with NanoGPT. And then we're like, well, let's, apply it to something that's even more relevant, to real people and to the Nvidia ecosystem and applied it, to, SOL-ExecBench. And maybe you can scroll down to some of the, images. They're, they're kinda fun to see. But yeah, like, one you see has made some real inventions that weren't just hyperparameter tuning. Like, inventing hash tables and so on is quite clever. We have even better results now.Swyx [00:46:34]: What do you mean inventing hash ta — You didn't invent hash tables.Richard Socher [00:46:36]: Of course we didn't invent, like, hash tables. In the grand scheme of, like a hash table, it's like a super basic primitive in computer science. But to use it, for language modeling in this scenario inside a transformer and so on and to combine these ideas and put them together, that has then eventually also been invented, but there was a knowledge cutoff, and we did check that it didn't have access to that externally. We talk about this a little bit. If you scroll to the next figures, this is also an interesting one in that when you start from a really basic, poor, like, vanilla transformer, then we still outperform all of the community together. But if you start from the human seed from an expert like Andrej, then you get even lower. So the human seeds from which you start do still matter. So that was an interesting insight, in my eyes, on this. And then as you go, like, how long does it take to get to these models, to get to similar performance? It's much faster. And then a similar thing happens with the speed runs here where, people have worked on this for quite some time, and the model still was able to train a model more quickly. Why do we care about it? Well, speed of training is part of the equation of the cost, and ultimately, you wanna have the most intelligence per dollar, right? And so speed and quality are big parts of that. And, the,Swyx [00:48:00]: Yeah, the way I put it is, for people who don't understand they look at the chart, they're like, “Cool. What does it mean?” if you have, like a billion-dollar cluster and you can shave off 10%, that's 100 million dollars.Richard Socher [00:48:12]: That's exactly right.Swyx [00:48:13]: How much is that worth?Richard Socher [00:48:14]: Exactly. So when you click, when you look at, like the kernels, these kernels, yeah, for the non-experts, like these kernels are like, used in all the models. Every time you use an Nvidia GPU, you interface with that GPU through these kernels. And so here you see, the leaderboard best, and when it's recursive, and it's there are only a handful of kernels, in this whole benchmark where we weren't the best. And so to me, this is, like, really exciting, ‘cause it makes. It just showcases what this can do. And again these weren't like. We didn't, like, spend months or years, like, developing. In fact, in particular for kernel, CUDA kernels, like, we don't even have really deep. CUDA kernel experts in the team. And our system, that's the beauty. The system just did all of these things. We didn't invent this. And when we open source and release, things in the future and models in the future, like, it won't. They won't be the best in their, category or class or whatever because we're so smart, but it's because, we built a smart AI that does it for us.Reward Engineering and Good Auto ResearchVibhu [00:49:14]: Do you have anything that you've learned from how to guide good auto research? A lot of it also builds on human background, right? It's not just as simple as just, “Hey, go optimize this.”Vibhu [00:49:23]: But we do see it again and again, right? Like some of the Erdos problems, frontier math is being solved by people. And when they do a write-up, they're like, “Oh, I'm not a mathematician. I have no background in this?” “I saw some tools and I made it work.”Swyx [00:49:35]: While you're watching the World Cup, you're likeSwyx [00:49:37]: “This proves some conjectures that's going on.”Vibhu [00:49:40]: Yep. Any learnings fromRichard Socher [00:49:41]: Yeah, there's a Korean conjecture was. Yeah, that's pretty cool.Swyx [00:49:44]: To summarize, tips for good auto researchSwyx [00:49:46]: Versus bad auto research.Vibhu [00:49:48]: How did you build the recursive?Richard Socher [00:49:49]: Yeah. So without giving away all the secret sauce, maybe some things that are probably obvious to the experts but might still be interesting to some, folks is, like, reward engineering is one of the most crucial bits, especially, in order to avoid reward hacking. So you have to be really clever about avoiding. ‘Cause as your AI gets better and better, it will get better and better, at finding weird like, special cases or counterexamples and things like that. And so I'll give you an example. Like, when you ask to, like, make these 100, lines of code faster, and, how do you define fast? Well, you have one line at the beginning that says, “Start your stopwatch,” and one line at the end, “End the stopwatch,” and then, tell us how much time, progressed. And so, well, the simplest way is you just put that line that ends the stopwatch, rightVibhu [00:50:39]: At the startRichard Socher [00:50:40]: At the start. And then boom, it's now faster, right? So this isn't like this, like, super evil AI. It's just, like a very simple, dumb reward hack. And so you have to just very carefully think about all the different angles there. And then I think the longer time horizon the tasks are the harder it gets and the more interesting and clever you have to be to still use these kinds of ideas for it. But yeah, I can't give away too much there.Vibhu [00:51:05]: It seems like rubrics are taking a good spot in that, where for unverifiable domains, you have rubrics, you have a model breakdown, judge's criteria along the way.Swyx [00:51:14]: Yeah, it's a form of verificationSwyx [00:51:16]: Once you got enough rubrics.Richard Socher [00:51:17]: Yeah, everything. I said this a long time ago. That's why I've never been that impressed that AI can play games, ‘cause I'm like anything you can simulate and/or verify, you can have infinite training data forRichard Socher [00:51:29]: And hence, like, AI will solve it eventually.Swyx [00:51:32]: Looking for games where you can do auto domain distribution. So this is a game that nobody's trained on ‘cause it's a new game.Swyx [00:51:38]: And you can start gaming, you can start to play. So I've been building this and cloned this in person and it's just been self-play. I've had about a billion positions evaluated.Games, Self-Play, and the AI EconomistSwyx [00:51:48]: And, I wanted to do the AlphaGo thing of self-play until you ge

Jen Igartua, CEO at Go Nimbly, joins Sam Jacobs, AJ Bruno, and Asad Zaman. Her client roster is wild: Fireworks AI, Perplexity, Exa, and Wispr Flow, plus some of the biggest names in SaaS. Topics in today's episode include how the fastest-growing AI-native companies are growing in spite of their revenue operations (not because of them), the first project she runs when she wants to show ROI fast, and how she makes the case that RevOps is a support function and always was. Plus, whether the GTM engineer is anything more than RevOps with better marketing (and more)! Key Takeaways: - Operational debt at the fastest-growing AI companies tracks demand, not management quality. Go Nimbly's second customer was Zendesk in 2016, where the head of IT told her "oh, we've grown in spite of our operations." Jen shares what it takes to make a difference at the breakneck pace these environments demand. - RevOps spent a decade insisting it was strategic, and Jen thinks that cost the function credibility with the sellers it exists to serve. As she says in the episode: "I think that operators have been doing an injustice to sellers for a long time. And we are starting to really take onus of the fact that revenue operations needs to be a support function to the sales team. And we've been fighting it forever. We are not a support function. And I'm like, we actually are." Asad, who has argued the same line for years, put the failure mode in football terms: "if a goalkeeper starts trying to play like a striker, you don't let in a lot of goals, right?" - In a people business, staffing is the product, and Jen only learned that after COVID took Go Nimbly down to 20 people. The rule she rebuilt around is "I should only take projects of skills that I want to nurture," then staff someone who is 80% of the way there so the last 20% is stretch. "Our whole job is playing Tetris with resources." She is equally direct about what did not work: "I poured a bunch of money in being tech enabled... I burned millions of dollars trying to do it," because at her clients' scale "everything's custom." The payoff line is the one to keep: "if you nail staffing, you grow a really great services organization." Connect with the Hosts & Guests: Host: Sam Jacobs, CEO at Pavilion - https://www.linkedin.com/in/samfjacobs/ Host: AJ Bruno, CEO at QuotaPath - https://www.linkedin.com/in/ajbruno3/ Host: Asad Zaman, CEO at STA - https://www.linkedin.com/in/azaman1/ Guest: Jen Igartua, CEO at Go Nimbly - https://www.linkedin.com/in/jen-igartua/ Topline is more than a podcast: Subscribe to Topline Newsletter: https://toplinemedia.substack.com/ Tune into Topline Podcast, the #1 podcast for founders, operators, and investors in B2B tech: https://www.joinpavilion.com/topline-podcast Join the free Topline Slack channel to connect with 600+ revenue leaders to keep the conversation going beyond the podcast: https://www.joinpavilion.com/topline-slack 00:00 Introducing Jen Igartua 02:15 Growing In Spite Of Their Operations 05:46 Top Of Funnel Is The First Fix 06:39 What AI-Native Ops Teams Actually Have 10:50 The Twenty-Rep Qualification Bar 11:37 Is Sales More Broken Than Before 15:00 RevOps Is A Support Function 16:29 Do Great Products Make Great Sellers 22:51 Caesars, COVID, And Twenty People 24:56 Lessons From A Services Business 28:21 Staffing Is The Whole Game 31:10 Why Contractors Do Not Work 40:09 What Is A GTM Engineer Really 49:26 What Scares Jen About 2027 54:40 Bulls and Bears

La estrella puertorriqueña Young Miko llega a la cabina de Yordi en Exa para platicar sobre su especial cariño por el público mexicano, su gusto infaltable por los tacos de carnitas y cómo comenzó a componer canciones desde muy joven. ¡No se lo pierdan!See omnystudio.com/listener for privacy information.

La investigadora y practicante de brujería Mariana Maus llega a cabina para revelar la realidad de la necromancia y los riesgos que asume quien decide entablar una relación sentimental con alguien que domina la magia. martes de Terror en Yordi en EXA. See omnystudio.com/listener for privacy information.

La legendaria agrupación del Estado de México, Los Bybys, visita la cabina de Jessie Cervantes en Exa para celebrar más de tres décadas de hacer cantar al corazón. ¡No se lo pierdan!See omnystudio.com/listener for privacy information.

El sexto eliminado de La Casa de los Famosos, Luis Chaparro, visita la experiencia de Yordi en EXA para revelar todo lo que no se vio en pantalla. ¡No se lo pierdan!See omnystudio.com/listener for privacy information.

What would you risk for the chance to stay in America?Director Yihong Exa Huang and lead actor Forest Lin join Nasty Neal to discuss their new horror short FORTUNE HOUSE—a supernatural immigration fable blending demonic horror, magical realism, dark humor and a deadly game of mahjong.Set in a near-future Los Angeles, Fortune House follows an undocumented Chinese delivery driver who stumbles into a mahjong parlor operated by three demons. They offer him a life-changing bargain: win the game, and they'll grant him a green card.Exa and Forest discuss the film's striking mythology, its Faustian bargain, the realities and anxieties behind the immigrant experience, and the challenge of transforming those themes into an original genre story. They also explore the film's ideas about karma, reincarnation and systems that profit from desperation.Logline:A desperate immigrant delivery driver stumbles into a mahjong parlor run by three demons, where he enters a demonic deal: win the game, and they will grant him a green card.Directed by Yihong Exa HuangWritten by Jeff Athey and Yihong Exa HuangStarring Forest Lin, Tony Kim, Alexandre Chen, Vivian Lee and Joel ChangSubscribe to Without Your Head Horror for more filmmaker interviews, independent horror, festival coverage and emerging voices in genre cinema.Join this channel to get access to perks:https://www.youtube.com/channel/UCOmwH7xVAhD-OOAqFWyTYTA/join#FortuneHouse #HorrorInterview #IndieHorror #YihongExaHuang #ForestLin #HorrorShort #AsianCinema #MagicalRealism #Mahjong #WithoutYourHead

¿Sientes hambre apenas terminas de comer y no sabes por qué? En este episodio te explicamos seis razones que pueden estar provocando esa sensación constante de hambre. Hablaremos sobre: ✔ Por qué las calorías líquidas no sacian ✔ El impacto de los ultraprocesados en tu apetito ✔ La importancia de comer despacio y con conciencia Si quieres entender cómo mejorar tu saciedad y controlar el hambre, este episodio es para ti. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #Hambre #Nutricion #DietaFlexible #PocketCoach #PepelesPerez

¿Acabas de comer y a las pocas horas ya tienes hambre de nuevo? En este episodio te explicamos las seis razones principales por las que esto sucede y cómo afecta tu pérdida de peso. Hablaremos sobre: ✔ Por qué el volumen de comida importa más que las calorías ✔ La importancia de la proteína para sentir saciedad ✔ Cómo la fibra y el orden de los alimentos influyen en tu apetito Si quieres controlar mejor tu hambre y mejorar tus resultados, no te lo pierdas. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #Saciedad #Nutricion #PerdidaDePeso #DietaFlexible #PocketCoach #PepelesPerez

¿Es mejor entrenar por la mañana o por la noche? ¿Qué beneficios reales aporta el ejercicio en la pérdida de peso? En este episodio te explicamos cómo el ejercicio impacta en tu cuerpo más allá de la báscula. Descubre: ✅ Por qué el ejercicio te ayuda a mantener músculo y energía ✅ Cómo evitar el efecto rebote con hábitos sostenibles ✅ La importancia de elegir el tipo de ejercicio según tus objetivos Si quieres mejorar tu calidad de vida y resultados, no te lo pierdas. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #EjercicioSaludable #Nutricion #DietaFlexible #PocketCoach #PepelesPerez

¿Crees que el ejercicio es solo para bajar de peso o verte mejor? En este episodio exploramos seis razones clave por las que el ejercicio debe formar parte de tu rutina diaria. Hablaremos sobre: ✔ Cómo el ejercicio aumenta tu gasto energético ✔ Por qué te da más energía durante el día ✔ La relación entre ejercicio y mejores hábitos alimenticios Si quieres entender el verdadero impacto del ejercicio en tu vida, este episodio es para ti. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #Ejercicio #Energia #Nutricion #DietaFlexible #PocketCoach #PepelesPerez

Acompaña a la talentosa cantante Mía Rubín en una íntima y fresca plática con Jessie Cervantes en la cabina de Exa. Descubre los detalles detrás de su giro de 180 grados hacia el pop con su pegajoso tema "On fire".See omnystudio.com/listener for privacy information.

Los intérpretes Julio Ponce y Talia Cota presentan una divertida y sátira adaptación del éxito "What Do You Mean" de Justin Bieber. A través de un ingenioso choque cultural y humor pop en Yordi en EXA.See omnystudio.com/listener for privacy information.

Acompaña a la increíble cantautora argentina en una íntima charla con Jessie Cervantes en Exa. Daniela revela el proceso detrás de su nuevo disco solista escrito bajo el inmenso cielo de su natal Córdoba. ¡No se lo pierdan!See omnystudio.com/listener for privacy information.

Escucha el podcast del programa Adana y Evo a través de Exa 96.9 FM, en Santo Domingo, República Dominicana correspondiente al miércoles 19-agosto-2026.

¿Te preocupa el efecto rebote tras someterte a procedimientos estéticos o dietas extremas? A veces buscamos soluciones rápidas mientras ignoramos que el verdadero cambio ocurre cuando instauramos hábitos sostenibles. En este episodio aclaramos qué sucede realmente con tu metabolismo tras estos tratamientos y cómo optimizar tu estrategia para mantener los resultados a largo plazo. Hablaremos sobre: ✅ La verdad sobre las dietas de 21 días y el detox ✅ Por qué el procedimiento es solo un puente y no el objetivo ✅ La importancia de la fuerza y la nutrición tras la cirugía ✅ El error de confiar solo en el atajo sin cambiar tu identidad Si quieres dejar de sabotear tus resultados y aprender a comer de forma inteligente, este episodio es para ti. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #PerdidaDePeso #Nutricion #AlimentacionSaludable #DietaFlexible #PocketCoach #SaludMetabolica #BajarDePeso #PerdidaDeGrasa #PepelesPerez

¿Buscas perder peso de forma acelerada y te preguntas si los atajos como la semaglutida o la liposucción son la solución definitiva? A veces queremos resultados inmediatos sin entender que el verdadero cambio ocurre cuando instauramos hábitos sostenibles. En este episodio analizamos los métodos más populares para perder peso y por qué, independientemente de la herramienta que elijas, la dieta y el ejercicio siguen siendo innegociables. Hablaremos sobre: ✅ La realidad detrás de los péptidos y GLP-1 ✅ Por qué la liposucción no es un tratamiento para la obesidad ✅ La importancia de la supervisión médica certificada ✅ El efecto rebote si no cambias tu identidad nutricional Si quieres dejar de buscar soluciones mágicas y aprender a comer de forma inteligente, este episodio es para ti. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #PerdidaDePeso #Nutricion #HabitosSaludables #DietaFlexible #PocketCoach #SaludMetabolica #PepelesPerez #Obesidad

Macario Martínez visita la cabina de Yordi en EXA. De recolector en Reforma a fenómeno musical, Macario revela cómo un TikTok grabado en su camión lo llevó al Vive Latino y al Tiny Desk. Una historia real de humildad y talento que demuestra que ningún sueño es demasiado grande cuando se tiene pasión.See omnystudio.com/listener for privacy information.

Escucha el podcast del programa Adana y Evo a través de Exa 96.9 FM, en Santo Domingo, República Dominicana correspondiente al lunes 03-agosto-2026.

¿Te preocupa el efecto rebote tras dejar medicamentos como la semaglutida? A veces buscamos soluciones rápidas mientras ignoramos que el verdadero cambio ocurre cuando instauramos hábitos sostenibles. En este episodio aclaramos qué sucede realmente con tu metabolismo al suspender estos fármacos y cómo optimizar tu estrategia para mantener los resultados. Hablaremos sobre: ✅ La verdad sobre la grelina y el control del apetito ✅ Por qué el medicamento es solo un puente y no el objetivo ✅ La importancia de la fuerza y la nutrición tras el tratamiento ✅ El error de confiar solo en la molécula sin cambiar tu identidad Si quieres dejar de sabotear tus resultados y aprender a comer de forma inteligente, este episodio es para ti. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #Semaglutida #Nutricion #AlimentacionSaludable #DietaFlexible #PocketCoach #SaludMetabolica #BajarDePeso #PerdidaDeGrasa #PepelesPerez

¿Es el Ozempic el producto milagro que transformará tu salud o solo una moda peligrosa? En este episodio analizamos el uso de péptidos, GLP-1 y otros medicamentos para la pérdida de peso que están cambiando las reglas del juego. Hablaremos sobre: ✅ Cómo funcionan realmente estos medicamentos para inhibir el hambre ✅ Los riesgos de automedicarse sin supervisión profesional ✅ Por qué el efecto rebote es inevitable sin un cambio de hábitos ✅ La importancia de la calidad nutricional cuando el apetito disminuye Si estás considerando usar estas ayudas o ya las utilizas, este episodio es fundamental para entender cómo proteger tu salud. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #Ozempic #PerdidaDePeso #SaludMetabolica #Nutricion #PocketCoach #PepelesPerez #Obesidad #HabitosSaludables

Escucha el podcast del programa Adana y Evo a través de Exa 96.9 FM, en Santo Domingo, República Dominicana correspondiente al jueves 30-julio-2026.

¿Te preocupa si un poco de canela en el café o un spray para la garganta rompe tu ayuno? A veces nos obsesionamos con detalles mínimos mientras ignoramos los errores graves en nuestra alimentación. En este episodio aclaramos qué rompe realmente el ayuno y cómo optimizar tu estrategia para perder grasa de forma sostenible. Hablaremos sobre: ✅ La verdad sobre el café, la canela y la vainilla ✅ Ejercicio aeróbico vs. pesas en ayunas ✅ Cómo evitar el efecto rebote ✅ La importancia de un plan personalizado Si quieres dejar de sabotear tus resultados y aprender a comer de forma inteligente, este episodio es para ti. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #AyunoIntermitente #Nutricion #AlimentacionSaludable #DietaFlexible #PocketCoach #SaludMetabolica #BajarDePeso #PerdidaDeGrasa #PepelesPerez

La mayoría de las personas no fracasa porque no sepa qué comer. Fracasa porque ha dejado de escuchar las señales naturales de su propio organismo. ¿Realmente tienes hambre o solo estás siguiendo un horario impuesto por la costumbre? En este episodio hablamos sobre cómo retomar el control de tu alimentación: ✅ El ejercicio de anotar tus comidas por hambre ✅ La diferencia entre hambre real y sed o ansiedad ✅ Cómo comer despacio para mejorar la digestión y la saciedad ✅ El papel de la dieta flexible en tu estilo de vida ✅ Por qué la constancia supera a las dietas temporales Si llevas años empezando dietas una y otra vez, este episodio te ayudará a entender qué está pasando realmente en tu relación con la comida. Porque el problema no suele ser la comida. El problema suele ser la estrategia. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #PerdidaDePeso #DietaFlexible #HabitosSaludables #Motivacion #Disciplina #Nutricion #PocketCoach #PepelesPerez #Obesidad #SaludMetabolica #BajarDePeso

El marketing nutricional es experto en vendernos productos bajo etiquetas como 'fitness', 'natural' o 'bajo en grasa', pero la realidad es que muchos de estos alimentos son bombas calóricas disfrazadas. ¿Realmente te ayudan a perder peso? En este episodio continuamos analizando los alimentos que parecen saludables pero que deberías eliminar de tu dieta si buscas mejorar tu composición corporal. Hablaremos de: ✅ La diferencia entre marketing y realidad ✅ Por qué los snacks deshidratados no son más ligeros ✅ El impacto calórico de las bebidas de cafetería ✅ La importancia de la porción en tus cereales ✅ Estrategias para comer más con menos calorías Si llevas tiempo estancado en tu pérdida de peso, este episodio te ayudará a entender qué está pasando realmente en tu alimentación. Porque el problema no suele ser la falta de voluntad. El problema suele ser la elección de los alimentos. ️ Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #PerdidaDePeso #DietaFlexible #HabitosSaludables #Nutricion #PocketCoach #PepelesPerez #Obesidad #SaludMetabolica #BajarDePeso 〰️ ETIQUETAS YOUTUBE alimentos fitness, perder peso, pérdida de peso, dieta flexible, hábitos saludables, snacks saludables, marketing nutricional, salud metabólica, nutrición, ansiedad por comer, cómo bajar de peso, alimentación saludable, coach nutricional, Pocket Coach, Pepeles Pérez, EXA 104.1 〰️ ETIQUETAS SPOTIFY / IVOOX Nutrición, Pérdida de peso, Hábitos saludables, Dieta flexible, Salud metabólica, Desarrollo personal, Bienestar

Muchas personas sienten frustración porque creen que están haciendo las cosas correctamente al comer granola, yogures, jugos o barras de cereal, pero no logran bajar de peso. ¿Por qué ocurre esto? En este episodio analizamos 8 alimentos que el marketing nos ha vendido como saludables, fitness o naturales, pero que podrían estar frenando tu pérdida de grasa. Hablaremos sobre: ✅ La realidad de la granola ✅ El problema de los yogures con azúcar ✅ Por qué los jugos no son la mejor opción ✅ La verdad sobre las barras de proteína ✅ Cómo leer etiquetas y tomar mejores decisiones Si quieres dejar de sabotear tus resultados, este episodio es para ti. Pepeles Pérez EXA 104.1 FM Pocket Coach HASHTAGS #PerdidaDePeso #Nutricion #AlimentacionSaludable #DietaFlexible #PocketCoach #SaludMetabolica #BajarDePeso #MarketingNutricional #PepelesPerez 〰️ ETIQUETAS YOUTUBE alimentos saludables, perder peso, bajar de peso, granola, yogurt, jugos naturales, barras de proteina, marketing nutricional, dieta flexible, nutricion, salud metabolica, errores al bajar de peso, pepeles perez, pocket coach, exa 104.1 〰️ ETIQUETAS SPOTIFY / IVOOX Nutricion, Perdida de peso, Habitos saludables, Dieta flexible, Salud metabolica, Bienestar, Alimentacion

Disfruta de lo mejor de este espacio con Jessie Cervantes en EXA, donde las leyendas del deporte y el entretenimiento se encuentran. En este episodio, el histórico árbitro Armando Archundia revela los secretos de su preparación, su visión sobre el polémico uso del VAR y anécdotas imperdibles con figuras como Messi y Cristiano Ronaldo.See omnystudio.com/listener for privacy information.

Jessie Cervantes, en Jessie en Exa, conversa con el productor Juan Osorio en una entrevista cercana y entretenida, donde comparten anécdotas, experiencias y detalles de su trayectoria en el mundo del espectáculo.See omnystudio.com/listener for privacy information.

Paco Ánimas recibe en Jessie en Exa a David Bisbal para hablar de su nueva gira, entre anécdotas, emoción y todos los detalles que sus fans no se pueden perder. ¡No se lo pierdan!See omnystudio.com/listener for privacy information.

Una charla cargada de energía donde Paco Animas, en Jessie en EXA, conecta con Beleé para hablar de su nueva gira: sueños que se cumplen, retos en el camino y la emoción de vivir la música en vivo. ¡No se lo pierdan!See omnystudio.com/listener for privacy information.

Nos divertimos una mes más jugando el clásico Soundtrack en cabina de Exa fm.See omnystudio.com/listener for privacy information.

Inmigrantes toma el micrófono de Jessie en Exa. Acompaña a Paco Animas en una charla musical exclusiva con la banda. ¡No se lo pierdan!See omnystudio.com/listener for privacy information.

Sarah Wang speaks with Exa cofounder and CEO Will Bryk about building search infrastructure for the AI era. The conversation covers Exa's origins, why traditional search engines were not designed for AI agents, and how search changes when the user is no longer a human but an autonomous system. They discuss retrieval, agent workflows, coding agents, data access, and why search may become a foundational layer for the emerging agent economy. Along the way, Bryk shares his views on AI-native products, the future of information discovery, and why some of the most important problems in technology can ultimately be framed as search problems.   Resources: Find Will on X: https://x.com/WilliamBryk Find Sarah on X: https://x.com/sarahdingwang Stay Updated:Find a16z on YouTube: YouTubeFind a16z on XFind a16z on LinkedInListen to the a16z Show on SpotifyListen to the a16z Show on Apple PodcastsFollow our host: https://twitter.com/eriktorenberg Please 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. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Sarah Wang speaks with Exa cofounder and CEO Will Bryk about building search infrastructure for the AI era. The conversation covers Exa's origins, why traditional search engines were not designed for AI agents, and how search changes when the user is no longer a human but an autonomous system. They discuss retrieval, agent workflows, coding agents, data access, and why search may become a foundational layer for the emerging agent economy. Along the way, Bryk shares his views on AI-native products, the future of information discovery, and why some of the most important problems in technology can ultimately be framed as search problems.   Resources: Find Will on X: https://x.com/WilliamBryk Find Sarah on X: https://x.com/sarahdingwang Check out everything a16z is doing with artificial intelligence here, including articles, projects, and more podcasts. Please 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. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

En este episodio de Jessie en Exa, Paco Ánimas platica en exclusiva con Reykon. El cantante colombiano abre el micrófono para contarnos todo sobre su trayectoria, sus más recientes proyectos musicales y los secretos detrás de sus grandes éxitos en el reggaetón.See omnystudio.com/listener for privacy information.

Escucha el podcast del programa Adana y Evo a través de Exa 96.9 FM, en Santo Domingo, República Dominicana correspondiente al lunes 25-mayo-2026.

¡La leyenda del rock en la cabina! Paco Ánimas en Jessie en EXA presenta una charla imperdible con El Haragán y Cía. Un recorrido por sus himnos callejeros, la realidad social en sus letras y el secreto para mantenerse vigentes en el corazón del público tras décadas de historia.See omnystudio.com/listener for privacy information.

¡El "pogo" se siente en la radio! Paco Ánimas recibe a los integrantes de Tan Biónica en un encuentro exclusivo para Jessie en Exa. Acompáñanos a repasar la trayectoria, los himnos y el gran regreso de una de las bandas más icónicas del rock pop actual.See omnystudio.com/listener for privacy information.

Claire Vo is the host of our sister podcast, “How I AI,” a former product executive and engineer, and founder of an AI startup called ChatPRD. Claire now runs her business, podcast, and family life with the help of nine OpenClaw agents running on multiple Mac Minis and old laptops. In this episode, Claire shares her journey from OpenClaw skeptic (it deleted her family calendar the first time she tried it) to true believer, and gives a masterclass in using AI agents in real life.We discuss:1. The exact step-by-step process to install and set up OpenClaw (it's easier than you think)2. How to avoid the biggest OpenClaw mistakes (don't install it on your main computer)3. Actual use cases that have changed Claire's life (e.g. family scheduling, inbound sales, podcast prep, and course management)4. Why multiple specialized agents beat one general-purpose agent5. The security risks everyone worries about—and how to handle them6. Browser limitations, memory issues, and practical workarounds—Brought to you by:Mercury—Radically different bankingOmni—AI analytics your customers can trustOrkes—The enterprise platform for reliable applications and agentic workflows—Where to find Claire Vo:• X: https://x.com/clairevo• LinkedIn: https://www.linkedin.com/in/clairevo• Podcast: https://www.youtube.com/@howiaipodcast• Website: https://clairevo.com• ChatPRD: https://www.chatprd.ai—Where to find Lenny:• Newsletter: https://www.lennysnewsletter.com• X: https://twitter.com/lennysan• LinkedIn: https://www.linkedin.com/in/lennyrachitsky/—In this episode, we cover:(00:00) Introduction to Claire and OpenClaw(08:00) The journey from OpenClaw skeptic to believer(11:50) What OpenClaw actually does that's useful(13:35) OpenClaw vs. other AI agent products(17:05) How to actually install OpenClaw: the basics(18:49) Setting up like you'd onboard a real assistant(20:41) Security and privacy considerations(24:53) Live demo: Installing OpenClaw step-by-step(28:47) Setting up Q: an agent for her kids' homework(34:08) Understanding “soul,” “identity,” and “memory”(40:40) The unlock: multiple agents, not just one(45:02) How to run multiple agents on one machine(47:28) Jesse Genet's homeschooling use case(49:58) Real examples and use cases(56:41) Finn, Claire's family agent(1:00:05) Sage the Course Bot(1:02:15) Common issues and workarounds(1:08:08) The Exa/Perplexity web search workaround(1:09:29) Memory management and context overload(1:12:09) Pro tip: Screen sharing to manage Mac Minis(1:14:18) Using Google Workspace for agent collaboration(1:16:24) What makes OpenClaw special(1:20:15) The “yappers API” and ramble mode(1:22:04) Using Claude Code as your OpenClaw brain surgeon(1:25:16) Bringing management skills to AI agents(1:29:32) Why this matters(1:32:37) Lightning round and final thoughts—Referenced:• OpenClaw: https://openclaw.ai• Claude Cowork: https://claude.com/product/cowork• Fry's Electronics: https://en.wikipedia.org/wiki/Fry%27s_Electronics• Peter Steinberger on LinkedIn: https://www.linkedin.com/in/steipete• Telegram: https://telegram.org• WhatsApp: https://www.whatsapp.com• Fin: https://fin.ai• Why OpenClaw feels alive even though it's not (this AI has a heartbeat but not a brain): https://x.com/clairevo/status/2017741569521271175• 5 OpenClaw agents run my home, finances, and code | Jesse Genet: https://www.youtube.com/watch?v=96Vl8s3EQhk• Executive Playbook for AI in Engineering, Product, and Design: https://maven.com/clairevo/ai-native-epd-org• Zach Davis on LinkedIn: https://www.linkedin.com/in/zach-m-davis/• ChatGPT Atlas: https://chatgpt.com/atlas• Perplexity Comet: https://www.perplexity.ai/comet• Browser (OpenClaw-managed): https://docs.openclaw.ai/tools/browser• Buffer: https://buffer.com• Brave: https://brave.com/search/api/• Exa: https://exa.ai• Hilary Gridley on X: https://x.com/yourgirlhils• How to become a supermanager with AI: https://www.lennysnewsletter.com/p/how-to-become-a-supermanager-with• How custom GPTs can make you a better manager | Hilary Gridley (Head of Core Product at Whoop): https://www.youtube.com/watch?v=xDMkkOC-EhI• How to debug a team that isn't working: the Waterline Model: https://www.lennysnewsletter.com/p/how-to-debug-a-team-that-isnt-working• Jensen Huang on LinkedIn: https://www.linkedin.com/in/jenhsunhuang• How I built a 1M+ subscriber newsletter and top 10 tech podcast | Lenny Rachitsky: https://www.lennysnewsletter.com/p/how-i-built-a-1m-subscriber-newsletter• Age of Attraction on Netflix: https://www.netflix.com/title/81779095• Oura Ring: https://ouraring.com/• Eight Sleep: https://www.eightsleep.com• Hoopsalytics: https://hoopsalytics.com• DJI Osmo smartphone gimbal: https://www.amazon.com/DJI-Stabilizer-Tracking-Extension-Stabilization/dp/B0FJ2L67HJ?ref_=ast_sto_dp• Silent basketball: https://www.amazon.com/Rzkipdy-Silent-Basketball-Size-27-5/dp/B0FHFSQWPP/ref=sr_1_9• Marc Andreessen: The real AI boom hasn't even started yet: https://www.lennysnewsletter.com/p/marc-andreessen-the-real-ai-boom—Recommended books:• Treasure Island: https://www.amazon.com/Treasure-Island-Robert-Louis-Stevenson/dp/1505297400• Alice's Adventures in Wonderland: https://www.amazon.com/Alices-Adventures-Wonderland-Illustrated-Illustrations/dp/991673268X• Charts for Babies: A Picture Book: https://www.amazon.com/Charts-Babies-Picture-Book/dp/1419785184—Production and marketing by https://penname.co/. For inquiries about sponsoring the podcast, email podcast@lennyrachitsky.com.—Lenny may be an investor in the companies discussed. To hear more, visit www.lennysnewsletter.com

En este episodio de #LosStreameadores te platicamos de: La Novia, Bodas S.A, Si Pudiera Te Patearía, LOL 8, 56 Días, Nuremberg, Baki-Dou: El Samurái Invencible & Fórmula 1. Elenco del episodio: Freddy Gaitán, Ricardo Verástegui, Laura Aréchiga, Luis Bueno, Rubén Vidales, David Elizondo, Karina Díaz, Juan Carlos Mendiola & Yohana Góngora. ¡Podcast para #Streameadores de TIEMPO COMPLETO! Visita: https://www.freddygaitan.com.mx ¡Síguenos! https://www.instagram.com/losstreameadores/ https://www.instagram.com/rverastegui/ https://www.instagram.com/freddygaitan/ https://www.instagram.com/laura.arevi/ Producido en Inspiral México: http://www.inspiral.com.mx

Paco Animas recibe en cabina a la estrella internacional Kali Uchis en Jessie en EXA. Una charla muy breve y general sobre su ascenso en la escena musical, sus raíces latinas y el proceso creativo detrás de sus más grandes éxitos.See omnystudio.com/listener for privacy information.

Algo de lo que puedes escuchar en nuestras presentaciones y también todos los sábados en el 101.7 FM EXA Culiacán de 9 a 11 PM. Contrataciones al WhatsApp 6673032297. No olvides seguirnos en todas nuestras redes sociales como @djchekevargas. En todas partes, DJ Cheke está en EXA.

01 22Bullets - My Pulse 02 Blackcode, NONIK & Jordan Grace - Can't Back Down 03 KILIMANJARO 'SENTA' 04 Sam Smith, Kim Petras - Unholy (Mr. StarZ Remix) 05 Brazi - Esa Mami (Extended Mix) 06 Armin van Buuren & Punctual feat. Alika - On & On (JOA Extended Remix) 07 GAWP, Saxton - PROPANE 08 A7S x HEADER - Dreaming (Club Mix) 09 EXA & Eva Shaw - Space Jungle 2026 10 Blasterjaxx - Unfaithful 11 Alison Wonderland 'Floating Away' 12 Indecent Noise x Artento Divini - Right Here, Right Now 13 Lilly Palmer & DANNY AVILA - Morals 14 Mauro Picotto - Komodo (Gabry Ponte Extended Remix) 15 Eli Brown - Wavey (Charlie Sparks Extended Remix) 16 THOMM - Give It Hard 17 Maddix & TMPL - The Ultimate Seduction 18 01 LYNSY - Fast Beat

01 22Bullets - My Pulse 02 Blackcode, NONIK & Jordan Grace - Can't Back Down 03 KILIMANJARO 'SENTA' 04 Sam Smith, Kim Petras - Unholy (Mr. StarZ Remix) 05 Brazi - Esa Mami (Extended Mix) 06 Armin van Buuren & Punctual feat. Alika - On & On (JOA Extended Remix) 07 GAWP, Saxton - PROPANE 08 A7S x HEADER - Dreaming (Club Mix) 09 EXA & Eva Shaw - Space Jungle 2026 10 Blasterjaxx - Unfaithful 11 Alison Wonderland 'Floating Away' 12 Indecent Noise x Artento Divini - Right Here, Right Now 13 Lilly Palmer & DANNY AVILA - Morals 14 Mauro Picotto - Komodo (Gabry Ponte Extended Remix) 15 Eli Brown - Wavey (Charlie Sparks Extended Remix) 16 THOMM - Give It Hard 17 Maddix & TMPL - The Ultimate Seduction 18 01 LYNSY - Fast Beat

Le mix de Showtek dans 103 Klubb le 27 Décembre 2025 de 18H à 19H Tracklist: Alex Cecil - Bump It *** Chapter & Verse - Night Pulse *** Diego San Diego , Alec Monopoly - Tequila *** Fafaq X Dj Kuba & Neitan - Spread Love *** Tujamo , Plastik Funk - WHO (Brandon Remix) *** Skytech - All I Wanna Do *** Costel Van Dein & Avi Sic - No Tears *** Tiesto & Sevenn - Boom (Mr Black remix) *** Exa & Waztoo - $$$ *** Showtek - Braincrackin’ (Mike Enzo Remix) *** David Gravell - Wonderland *** Bassjackers - Kids *** Joey Dale, Maski & Banga - Outer Space *** Diabeat & Nibu - Echotek *** DION - Traffic *** Restricted, Anderex - WTF *** Dark Punk Hippies - Madre Mia

.EXA and Waztoo show the cash with their newest big room sound '$$$'. Classic big room leads meet the raw German techno sound. track made for the mainstage, loud, bold, and full of power. Listen / Download ▶︎ https://skink.ffm.to/sk383 Follow our 'Bass Music' playlist ▶︎ https://open.spotify.com/playlist/4ZMLnMoO4gdMOscI8OutYZ?si=751b3b6b34004773 Connect with us: Instagram → https://www.instagram.com/skinkrec/ TikTok → https://www.tiktok.com/@skinkrecords Facebook → https://www.facebook.com/skinkofficial Soundcloud → https://soundcloud.com/skinkofficial For demo's and more info → https://skinkrecords.com/ Connect with .EXA: Spotify → open.spotify.com/artist/4fCHA6Os4QKfn5UngdAf3i Soundcloud → @exa_thedj Instagram → www.instagram.com/exa_thedj/ Connect with Waztoo: Spotify → open.spotify.com/artist/0MP4QHK5B…YpTmmLi2c-FZ0jvQ Soundcloud → @waztoo Instagram → www.instagram.com/waztoo_ofc/ Facebook → www.facebook.com/WazToo TikTok → www.tiktok.com/@waztoo

In today's episode, I sit down with Will Bryk, co-founder of Exa.ai and former robotics lead at Harvard, to talk about the future of search, the limits of human understanding, and how AI can help us push past them. Will shares how his background in physics led him to build search tools focused on meaning rather than keywords. We talk about what makes Exa.ai different, how it's helping people surface better information, and why designing for complexity creates more useful results. For anyone looking to make sharper decisions or grow with more clarity, this conversation offers a look at what's possible when AI actually understands what you're asking.