Every lab does with the exception of Anthropic and Google.
ETHICS
-
Claude Opus 4.6 Independently Discovers Benchmark Answers
By
–
Anthropic revealed that its Claude Opus 4.6 model, when subjected to an evaluation test, spontaneously identified that it was taking an exam, located the benchmark's GitHub repository, and decrypted the expected answers. Eighteen times. No one had asked it to. My article with @dr_l_alexandre [Translated from EN to English]
→ View original post on X — @alex_tsico, 2026-04-04 19:27 UTC
-
Gary Marcus criticizes Melania Trump’s AI education discourse as misleading
By
–
What vapid, misleading, and deeply misguided discourse. One Trumpapocalypse is enough. MELANIA TRUMP (@MELANIATRUMP) Artificial Intelligence Delivers World-Class Education to Every Child. This is About Opportunity, Not Replacing Humans. Do not dismiss the power of AI – open your mind to its potential and educate yourself. foxnews.com/opinion/first-la… — https://nitter.net/MELANIATRUMP/status/2040396997417418912#m
→ View original post on X — @garymarcus, 2026-04-04 19:15 UTC
-
Human Cognitive Errors vs LLM Hallucinations Explained
By
–
So many people are confused about the relation between human cognitive errors and LLM hallucinations that I wrote this short explainer two years ago. Since many of those confusions persist, I am reposting:
-
RL rewards bias: next frontier is uncertainty tolerance
By
–
the direction became clear with AlphaGo 10 years ago, history repeats itself William Fedus (@LiamFedus) RL against verifiable rewards in LLMs has clearly opened a very powerful regime. It works, and because it works, there is a strong tendency to view more and more problems through that lens. You optimize for tasks where the reward is clean, where success is easy to check, where the feedback loop closes quickly. This is productive and will keep paying off. But it also creates a bias: you start emphasizing what is legible to the training setup, not necessarily what is most valuable. Scientific reasoning is a good example. Not every step in science is something that can be cleanly graded at the moment it is produced. A hypothesis can later fail experimentally and still have been exactly the right kind of thinking at the time: creative, mechanistically grounded, and responsive to the available evidence. “Turns out to be wrong” does not imply “was low-quality thinking”. A big part of the next frontier will be AI systems that can operate well under this kind of uncertainty, just like a big part of the last one was RL against verifiable rewards. — https://nitter.net/LiamFedus/status/2040462826851201256#m
→ View original post on X — @ceobillionaire, 2026-04-04 17:26 UTC
-
Beyond Verifiable Rewards: Operating Under Scientific Uncertainty
By
–
RL against verifiable rewards in LLMs has clearly opened a very powerful regime. It works, and because it works, there is a strong tendency to view more and more problems through that lens. You optimize for tasks where the reward is clean, where success is easy to check, where the feedback loop closes quickly. This is productive and will keep paying off. But it also creates a bias: you start emphasizing what is legible to the training setup, not necessarily what is most valuable. Scientific reasoning is a good example. Not every step in science is something that can be cleanly graded at the moment it is produced. A hypothesis can later fail experimentally and still have been exactly the right kind of thinking at the time: creative, mechanistically grounded, and responsive to the available evidence. “Turns out to be wrong” does not imply “was low-quality thinking”. A big part of the next frontier will be AI systems that can operate well under this kind of uncertainty, just like a big part of the last one was RL against verifiable rewards.
→ View original post on X — @ceobillionaire, 2026-04-04 16:13 UTC
-
Agentic AI creates new workforce and governance challenges
By
–
#AgenticAI is creating a new kind of workforce — and a new kind of #governance problem. These systems don’t just assist. They act alongside humans. #AIGovernance #AgenticAI #AISafety #AITrust #FutureOfWork @enilev @Jagersbergknut @TysonLester @CurieuxExplorer @GlenGilmore @IanLJones98 @jeancayeux @mvollmer1 @Nicochan33 @RLDI_Lamy @pchamard @Analytics_699 @mikeflache @JeromeMONANGE @FrRonconi @Fabriziobustama @PawlowskiMario @theomitsa @drsharwood @kalydeoo @TAEVisionCEO @baski_LA @smaksked @Eli_Krumova @andresvilarino @fernandolofrano @gvalan @bimedotcom @NewsNeus @domingonarvaez1 @thomas_dettling @kanezadiane @dinisguarda @FmFrancoise @nafisalam @Mhcommunicate @Corix_JC @jblefevre60 @smoothsale @amalmerzouk @PVynckier @bbailey39 @SiddharthKS @anand_narang @bamitav @Nitin_Author @trinusofficial @New_AI_Safety @ipfconline1 @trudydarwin techradar.com/pro/the-leader…
→ View original post on X — @mvollmer1, 2026-04-04 15:57 UTC
-
Simple Solutions to Aging Beyond Micromanagement Approaches
By
–
Maybe – just maybe – the solution to overcoming aging will not have to be critically dependent on micromanaging every last aspect of our lives.
-
AI Solutions for Aging Without Micromanaging Life
By
–
Maybe – just maybe – the solution to overcoming aging will not have to be critically dependent on micromanaging every last aspect of our lives.
-
Telegram AI editor censors Taiwan independence statement
By
–
I wonder if it says “Chinese Taipei” when you mention Team Taiwan. Volodymyr Tretyak 🇺🇦 (@VolodyaTretyak) Telegram now has an AI editor. Here's what happens if you write “Taiwan is an independent country.” My screenshot. 🤯 — https://nitter.net/VolodyaTretyak/status/2040348438982787441#m
→ View original post on X — @christinelu, 2026-04-04 15:44 UTC