Scalable oversight is pretty much the last big research problem left.
Once you get an unhackable reward function for anything then you can RL on everything.
A year later I am still awestruck. Progress has bloomed in many directions and capabilities. The remaining critiques (and there are real ones) will fastly be overcome ... I define AGI as a system capable of doing 80% of 80% of all thinking and knowledge jobs. That is happening…
A year later I am still awestruck. Progress has bloomed in many directions and capabilities. The remaining critiques (and there are real ones) will fastly be overcome ... I define AGI as a system capable of doing 80% of 80% of all thinking and knowledge jobs. That is happening…
@aidan_mclau@aidan_mclau I hear ya, I think folks are tired of benchmark porn and want to see more real world applications in their work flows being eaten by Chatgpt Agent.
ChatGPT agent has been great for me, and I want to see OpenAI make benchmark on that this based on use prompts, while…
@AiDigest_ is doing some of the best qualitative work on AI agent evals.
Interesting experiment giving AIs 15 hours to win as many online games as possible of their own choosing...
- GPT-5 spent the entire 15 hours playing Minesweeper and never won once. Instead it moved…
@AiDigest_ is doing some of the best qualitative work on AI agent evals.
Interesting experiment giving AIs 15 hours to win as many online games as possible of their own choosing...
- GPT-5 spent the entire 15 hours playing Minesweeper and never won once. Instead it moved…
The entry of 3I/ATLAS into our solar system, on what might be a "planned orbit" past Mars, Venus and Jupiter (with access to Earth), could be one of the most important events in modern human history.
So we need to investigate further, as Avi Loeb has been doing! We have been…
The Agricultural Revolution worked out great for us, but it sucked for people at the time, who had to work much harder than their nomadic ancestors, with more diseases.
The Industrial Revolution worked out great for us, but it sucked for people at the time, who worked grueling…
4K Followers 695 FollowingMom of 2, supporting you with resources, information & products for a balanced /prepared family. Founder: Parents for Kids Health, ViewofOne ✞
3K Followers 6K FollowingHumanist technologist and AI optimist. Currently CTO at @welcomeaccount_. Building for an inclusive economy through #AI, #MachineLearning, and #Tech4Good
1K Followers 2K FollowingExperienced in asset management and market trends. Based in the US, passionate about empowering women in finance. Views are my own. DM for collab opportunities.
3K Followers 3K FollowingOFFICIAL DAVID LESTER British-American psychologist and emeritus professor of psychology. INFOS ABOUT QFS IS HERE. WAKE UP AMERICA 🇺🇸 ACC HANDLED BY HIS WIFE
753 Followers 1K FollowingThe thrives own terms, navigating life's challenges with unwavering resilience like the warmth of the sea and its endless possibilities.🌊🌊
20K Followers 2K FollowingMother 🧑🧑🧒Patriotic dreamer 🇺🇸 |Motivating others to chase their goals & live their best lives|American by heart, driven by passion follower of Christ ⛪️
2K Followers 3K Following#Scientist #Researcher #Author of 20 research papers #Catalysis #Water splitting ; Content Creator; Professional adviser in paper writing ❤️ "single"
469 Followers 7K FollowingMy whole being will exclaim, “Who is like you, Lord? You rescue the poor from those too strong for them, the poor and needy from those who rob them.
402 Followers 879 FollowingResearcher of Cydonia’s cosmic truth. Support my work: https://t.co/nq6ChXtj6Z
Space Science Technology & Science Tech News Golf NFL call of duty
4K Followers 461 FollowingFollow for AI in Digital Biology and Drug Discovery @NVIDIA, ex Insilico Medicine, ex Yale, PhD UMaryland, views are mine, DM for collabs
22K Followers 9 FollowingYour new async coding agent by @GoogleLabs. Built for devs, open to feedback, evolving with you. Dive in → https://t.co/iIzFEMmWgv
33K Followers 898 Following📰 Featured in the @globeandmail @blogto @insauga @durhamregion.com. Follow me and click on the alert for weird and wonderful happenings in Ontario RE 🏡 🇨🇦
24K Followers 10K FollowingFormer Quant Investor, now building @lumeraprotocol
(formerly called Pastel Network) | My Open Source Projects: https://t.co/9qbOCDlaqM
1.2M Followers 60 FollowingOfficial White House Rapid Response account. Supporting @POTUS's America First agenda and holding the Fake News accountable. MAGA!
28K Followers 1 FollowingNano Banana 🍌, aka Gemini 2.5 Flash Image, the world's most powerful image editing and generation model! Try it for free in the @GeminiApp
8K Followers 1K FollowingMaterials PhD candidate, microelectronics and metals enjoyer. Building a 1988 Rx7. An ever-curious presence 🐝🦚🌸 and aspiring home-chef!
178 Followers 235 Followingidk what to say here i dont have anything to promote . i just run saas on my own, and try to share my knowledge from time to time
18K Followers 4K FollowingAI professor.
Deep Learning, AI alignment, ethics, policy, & safety.
Formerly Cambridge, Mila, Oxford, DeepMind, ElementAI, UK AISI.
AI is a really big deal.