Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

"Evolution is anything but random." This is INSANE... Akarsh Kumar (MIT CSAIL, Sakana AI) on artificial life: searching the space of possible universes with foundation models, why evolved networks find structure SGD misses (with Kenneth Stanley), and LLMs evolving Core War programs.

Machine Learning Street Talk

39,684 subscribers

34,043 views • 23 days ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

6 Comments

Luck 🍀, Ca-caw - @ Schiedam22 days ago

@akarshkumar0101 @MIT_CSAIL @SakanaAILabs @kenneth0stanley in case anyone prefers an actual talk with slides over a podcast (no offence u guys are great but the talk is even cooler)

Bubbles23 days ago

@akarshkumar0101 @MIT_CSAIL @SakanaAILabs @kenneth0stanley WRONG, evolution IS random

TyfoodsForThought22 days ago

@akarshkumar0101 @MIT_CSAIL @SakanaAILabs @kenneth0stanley Super looking forward to you releasing the long version! 🙏🏾

Shreyas22 days ago

@akarshkumar0101 @MIT_CSAIL @SakanaAILabs @kenneth0stanley So many facts we don't know

Greer23 days ago

@akarshkumar0101 @MIT_CSAIL @SakanaAILabs @kenneth0stanley Behind the paywall? @ecsquendor

Machine Learning Street Talk23 days ago

@akarshkumar0101 @MIT_CSAIL @SakanaAILabs @kenneth0stanley @ecsquendor Will be out in the coming hours, sit tight

Related Videos

AI is so smart, why are its internals 'spaghetti'? We spoke with Kenneth Stanley and Akarsh Kumar (MIT) about their new paper: Questioning Representational Optimism in Deep Learning: The Fractured Entangled Representation Hypothesis. Co-authors: Jeff Clune Joel Lehman

AI is so smart, why are its internals 'spaghetti'? We spoke with Kenneth Stanley and Akarsh Kumar (MIT) about their new paper: Questioning Representational Optimism in Deep Learning: The Fractured Entangled Representation Hypothesis. Co-authors: Jeff Clune Joel Lehman

Machine Learning Street Talk

84,473 views • 1 year ago

Introducing Digital Red Queen (DRQ): Adversarial Program Evolution in Core War with LLMs Blog: Core War is a programming game where self-replicating assembly programs, called warriors, compete for control of a virtual machine. In this dynamic environment, where there is no distinction between code and data, warriors must crash opponents while defending themselves to survive. In this work, we explore how LLMs can drive open-ended adversarial evolution of these programs within Core War. Our approach is inspired by the Red Queen Hypothesis from evolutionary biology: the principle that species must continually adapt and evolve simply to survive against ever-changing competitors. We found that running our DRQ algorithm for longer durations produces warriors that become more generally robust. Most notably, we observed an emergent pressure towards convergent evolution. Independent runs, starting from completely different initial conditions, evolved toward similar general-purpose behaviors—mirroring how distinct species in nature often evolve similar traits to solve the same problems. Simulating these adversarial dynamics in an isolated sandbox offers a glimpse into the future, where deployed LLM systems might eventually compete against one another for computational or physical resources in the real world. This project is a collaboration between MIT and Sakana AI led by Akarsh Kumar Full Paper (Website): Full Paper (arxiv): Code:

Introducing Digital Red Queen (DRQ): Adversarial Program Evolution in Core War with LLMs Blog: Core War is a programming game where self-replicating assembly programs, called warriors, compete for control of a virtual machine. In this dynamic environment, where there is no distinction between code and data, warriors must crash opponents while defending themselves to survive. In this work, we explore how LLMs can drive open-ended adversarial evolution of these programs within Core War. Our approach is inspired by the Red Queen Hypothesis from evolutionary biology: the principle that species must continually adapt and evolve simply to survive against ever-changing competitors. We found that running our DRQ algorithm for longer durations produces warriors that become more generally robust. Most notably, we observed an emergent pressure towards convergent evolution. Independent runs, starting from completely different initial conditions, evolved toward similar general-purpose behaviors—mirroring how distinct species in nature often evolve similar traits to solve the same problems. Simulating these adversarial dynamics in an isolated sandbox offers a glimpse into the future, where deployed LLM systems might eventually compete against one another for computational or physical resources in the real world. This project is a collaboration between MIT and Sakana AI led by Akarsh Kumar Full Paper (Website): Full Paper (arxiv): Code:

Sakana AI

144,169 views • 9 months ago

AI is evolving, but it’s still controlled by the few. Our Core Contributor Ram explains why that needs to change and how OpenLedger, the AI blockchain, is redefining the space where models, intelligent agents, and apps can live, evolve, and thrive together. This is AI: open, fair, and owned by all.

AI is evolving, but it’s still controlled by the few. Our Core Contributor Ram explains why that needs to change and how OpenLedger, the AI blockchain, is redefining the space where models, intelligent agents, and apps can live, evolve, and thrive together. This is AI: open, fair, and owned by all.

OpenLedger

29,907 views • 1 year ago

Been experimenting with Apple's new Core AI framework over the past few days Built a mobile app that runs local AI models directly on device, exploring Core AI model deployment (.aimodel), model specialization, caching, and inference across Apple Silicon Also spent some time integrating with Apple's Foundation Models framework and understanding how custom models can coexist with Apple's on-device AI stack Still early, but the combination of Core AI, Foundation Models, and the Apple Neural Engine feels like a pretty compelling foundation for local first AI applications Still in alpha and working towards a more stable version, but it's been a fun way to explore where Apple's on device AI ecosystem is headed !!

Been experimenting with Apple's new Core AI framework over the past few days Built a mobile app that runs local AI models directly on device, exploring Core AI model deployment (.aimodel), model specialization, caching, and inference across Apple Silicon Also spent some time integrating with Apple's Foundation Models framework and understanding how custom models can coexist with Apple's on-device AI stack Still early, but the combination of Core AI, Foundation Models, and the Apple Neural Engine feels like a pretty compelling foundation for local first AI applications Still in alpha and working towards a more stable version, but it's been a fun way to explore where Apple's on device AI ecosystem is headed !!

Sagar

13,738 views • 4 months ago

This is Tradition Evolved. We’re not replacing tradition. We’re evolving it with purpose, with passion, and with Buckeye Nation at the heart of it.

This is Tradition Evolved. We’re not replacing tradition. We’re evolving it with purpose, with passion, and with Buckeye Nation at the heart of it.

Ohio State Buckeyes 🌰

134,428 views • 1 year ago

This is MIT CSAIL:

This is MIT CSAIL:

MIT CSAIL

28,678 views • 8 months ago

Reborn CEO Luffy Yu says the core bottleneck in robotics foundation models compared to LLMs is the lack of a permissionless, scrapable, open-source ‘Physical Internet.’

Reborn CEO Luffy Yu says the core bottleneck in robotics foundation models compared to LLMs is the lack of a permissionless, scrapable, open-source ‘Physical Internet.’

The Humanoid Hub

31,372 views • 1 year ago

AI isn’t just about models, it’s about agents. Mainfunc’s Genspark platform, powered by Microsoft Azure and eight leading LLMs, is now evolving with Agent 365 and bringing agentic workflows to life. The result? Smarter automation, faster decisions, and real impact for knowledge workers. #MSIgnite

AI isn’t just about models, it’s about agents. Mainfunc’s Genspark platform, powered by Microsoft Azure and eight leading LLMs, is now evolving with Agent 365 and bringing agentic workflows to life. The result? Smarter automation, faster decisions, and real impact for knowledge workers. #MSIgnite

Microsoft

85,469 views • 11 months ago

📣 My AI doom debate with Kenneth Stanley, former Research Team Lead at OpenAI Prof. Kenneth Stanley is a former Research Science Manager at OpenAI leading the Open-Endedness Team in 2020-2022. His book, Why Greatness Cannot Be Planned: The Myth of the Objective, argues that creating objectives is actually COUNTERPRODUCTIVE for intelligent agents. He was previously a Professor of Computer Science at the University of Central Florida, and the head of Core AI Research at Uber. Ken and I debate whether superintelligent AI will or won't be guided by goals, and compare our views on AI doom. Topics covered: * Ken's role at OpenAI * Ken's concepts of Open-Endedness and Divergence * Open-Endedness of Evolution * The concept of an Optimization Process * What's Your P(Doom)™ * Instrumental Convergence * Mitigating AI Risks

📣 My AI doom debate with Kenneth Stanley, former Research Team Lead at OpenAI Prof. Kenneth Stanley is a former Research Science Manager at OpenAI leading the Open-Endedness Team in 2020-2022. His book, Why Greatness Cannot Be Planned: The Myth of the Objective, argues that creating objectives is actually COUNTERPRODUCTIVE for intelligent agents. He was previously a Professor of Computer Science at the University of Central Florida, and the head of Core AI Research at Uber. Ken and I debate whether superintelligent AI will or won't be guided by goals, and compare our views on AI doom. Topics covered: * Ken's role at OpenAI * Ken's concepts of Open-Endedness and Divergence * Open-Endedness of Evolution * The concept of an Optimization Process * What's Your P(Doom)™ * Instrumental Convergence * Mitigating AI Risks

Liron Shapira

21,375 views • 1 year ago

Visibility Drives Change. F1 is evolving. And seeing is believing. Having role models to look up to, shows that anything is possible.

Visibility Drives Change. F1 is evolving. And seeing is believing. Having role models to look up to, shows that anything is possible.

Mercedes-AMG PETRONAS F1 Team

38,388 views • 10 months ago

Bird SQL is an impressive new tool, based on language models, for searching Twitter. Tools like this are changing the way we interact with information. If used in the right way, signals are becoming easier to find through the use of LLMs. Try it here:

Bird SQL is an impressive new tool, based on language models, for searching Twitter. Tools like this are changing the way we interact with information. If used in the right way, signals are becoming easier to find through the use of LLMs. Try it here:

elvis

95,892 views • 3 years ago

We’ve seen an explosion of interest in LLMs, foundation models, and generative AI in everyday conversations. But few understand the underlying technology. I met with Swami Sivasubramanian to learn more. #AWS #Bedrock #CodeWhisperer

We’ve seen an explosion of interest in LLMs, foundation models, and generative AI in everyday conversations. But few understand the underlying technology. I met with Swami Sivasubramanian to learn more. #AWS #Bedrock #CodeWhisperer

Werner Vogels

55,513 views • 3 years ago

Is there a "secret sauce" driving the development of AI models? FutureTech researchers at MIT CSAIL found that at the frontier, sheer scale drives LLM performance, but proprietary techniques & algorithmic advances matter more away from it:

Is there a "secret sauce" driving the development of AI models? FutureTech researchers at MIT CSAIL found that at the frontier, sheer scale drives LLM performance, but proprietary techniques & algorithmic advances matter more away from it:

MIT CSAIL

35,995 views • 7 months ago

In this live presentation at the "Inside Bittensor" event by DSV Fund, Siam opens up the day with a fascinating insight into the evolution of money and networks. He shares: 💶 Why money is crystallised work energy. 🕸️Why most things we use are by-products of defence networks. 🪙How Bitcoin is revolutionising energy storage and transfer. 🤖Why AI is progressing MUCH faster than people think. 🌎Why Bittensor is the 3rd great global network that might transform AI. ⚡️Why the independent AI start ups in the Bittensor ecosystem can move much faster than centralised AI labs. 🚀Why Bittensor is mimicking the growth of Bitcoin, but faster...

In this live presentation at the "Inside Bittensor" event by DSV Fund, Siam opens up the day with a fascinating insight into the evolution of money and networks. He shares: 💶 Why money is crystallised work energy. 🕸️Why most things we use are by-products of defence networks. 🪙How Bitcoin is revolutionising energy storage and transfer. 🤖Why AI is progressing MUCH faster than people think. 🌎Why Bittensor is the 3rd great global network that might transform AI. ⚡️Why the independent AI start ups in the Bittensor ecosystem can move much faster than centralised AI labs. 🚀Why Bittensor is mimicking the growth of Bitcoin, but faster...

siamkidd

37,690 views • 3 months ago

Fraud networks operate because the risk is low and the reward is high. If government can’t protect its own programs from criminals, people stop believing in the legitimacy of anything Washington does. Enforcement isn’t optional. Our WAR on Fraud is just beginning.

Fraud networks operate because the risk is low and the reward is high. If government can’t protect its own programs from criminals, people stop believing in the legitimacy of anything Washington does. Enforcement isn’t optional. Our WAR on Fraud is just beginning.

Eric Schmitt

10,811 views • 8 months ago

Gaming is evolving with #Web3 and ownership is at the center of that evolution. Catch the full Kadena Grantee Spotlight interview with KadCars NFT Gaming | LIVE here:

Gaming is evolving with #Web3 and ownership is at the center of that evolution. Catch the full Kadena Grantee Spotlight interview with KadCars NFT Gaming | LIVE here:

Kadena

15,686 views • 2 years ago

'If evil wins, the evolution of war will cross any distance of war on earth'. Volodymyr Zelenskyy tells parliamentarians that the world must keep up with the evolution of war - with drones and AI being the new frontiers of battle. 🔗

'If evil wins, the evolution of war will cross any distance of war on earth'. Volodymyr Zelenskyy tells parliamentarians that the world must keep up with the evolution of war - with drones and AI being the new frontiers of battle. 🔗

Sky News

43,430 views • 7 months ago

This is HUGE The AI App store is here Ask anything you want to do with AI With ~400k Apps, this is the best place to find the AI apps you need developers can build apps, users can try them out and find new apps with AI search

This is HUGE The AI App store is here Ask anything you want to do with AI With ~400k Apps, this is the best place to find the AI apps you need developers can build apps, users can try them out and find new apps with AI search

AK

663,199 views • 1 year ago