正在加载视频...

视频加载失败

Chollet François Chollet : [paraphrase] "using NNs for discrete problems i.e. finding new prime numbers - is a terrible idea" with Kevin Ellis Zenna Tavares

23,254 次观看 • 1 年前 •via X (Twitter)

10 条评论

AGI Fire Alarm 的头像
AGI Fire Alarm1 年前

@fchollet @ellisk_kellis @ZennaTavares The manifold hypothesis may apply, to an extent, to algorithms that efficiently search for prime numbers

Jamesb 的头像
Jamesb1 年前

@fchollet @ellisk_kellis @ZennaTavares looking forward to this one! love the ideas behind dreamcoder!

{JSON: Huang} 的头像
{JSON: Huang}1 年前

@fchollet @ellisk_kellis @ZennaTavares Does this contradict the FrontierMath results in any way?

Venkat Madala 的头像
Venkat Madala1 年前

@fchollet @ellisk_kellis @ZennaTavares Totally agreed, let’s keep it prime

Tobias Holgersen 的头像
Tobias Holgersen1 年前

@fchollet @ellisk_kellis @ZennaTavares What is this from?

Machine Learning Street Talk 的头像
Machine Learning Street Talk1 年前

@fchollet @ellisk_kellis @ZennaTavares Unreleased episode

phi ARCHITECT 的头像
phi ARCHITECT1 年前

@fchollet @ellisk_kellis @ZennaTavares I think the problem is in trying to be reasonable about systems that can only hallucinate

Vlad Ciobanu 的头像
Vlad Ciobanu1 年前

@fchollet @ellisk_kellis @ZennaTavares yes, that's why o3 + a code interpreter is closer to a general intelligence than just o3 by itself

Santos L. Halper 的头像
Santos L. Halper1 年前

@fchollet @ellisk_kellis @ZennaTavares Matches with the fact that brains are not that great at finding primes

Uri Gil 的头像
Uri Gil1 年前

@fchollet @ellisk_kellis @ZennaTavares nobody ever said AGI would replace a calculator. intelligence benefit from and requires tools to be useful

相关视频

François Chollet (François Chollet) has spent years asking a different question than most of the AI world. Instead of scaling what already works, he’s trying to understand what intelligence actually is and how to build it from first principles. In this episode of the Lightcone Podcast, he traces that path from his early work on deep learning to the creation of the ARC Prize, and the launch of ARC V3, a new benchmark designed to measure something deeper than performance: the ability to learn, adapt, and reason efficiently in entirely new environments. He explains why today’s systems may be hitting limits, what recent breakthroughs really mean, and why reaching true general intelligence may require a fundamentally different approach. 00:00 - AGI by 2030? 00:31 - Introducing Ndea: A New Path Beyond Deep Learning 01:08 - A New ML Paradigm 01:30 - Replacing neural nets with compact symbolic programs 03:04 - Why Ndea Isn’t Competing With Coding Agents 05:20 - Why Everyone Might Be Wrong About Scaling LLMs 07:22 - Why Coding Agents Suddenly Work So Well 08:50 - The Limits of LLMs in Non-Verifiable Domains 10:48 - What AGI Actually Means (And Why Most Definitions Are Wrong) 13:30 - Why Deep Learning Hits a Wall 14:00 - ARC’s Origin Story 18:20 - ARC Benchmarks Explained: From V1 to V3 22:49 - The RL Loop Powering Coding Agents Today 27:03 - ARC-AGI V3: Measuring “Agentic Intelligence” 31:14 - Inside the ARC Game Studio 35:31 - Could AGI Fit in 10,000 Lines of Code? 44:01 - Building Ndea: From Idea to Compounding Research Stack 46:46 - The Future of ARC: Benchmarks That Evolve With AI 47:21 - Why There’s Still Huge Opportunity for New AI Paradigms 53:37 - How to Build a Breakout Open Source Project - Lessons From Keras 56:39 - Advice For How To Think About AI

Y Combinator

151,649 次观看 • 4 个月前