korrents

On the map

Tap a claim on the ring to put it at the centre.

← Many alignment-relevant properties, such as goal seeking and reward…

17 connected korrents · 15 moments from 18 Mar 2024 to 17 Nov 2026. Nearly all of them are about reinforcement learning.

Everything filed under AI alignment AI alignment Everything filed under AGI AGI Everything filed under AI agents AI agents Everything filed under OpenAI OpenAI Everything filed under LLMs LLMs Everything filed under neural networks neural networks Everything filed under benchmarks benchmarks Same subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subject Read this korrent: Many alignment-relevant properties, such as goal seeking and reward hacking, are plausibly introduced during RL training. Many alignment-relevant properties, suchas goal seeking and reward hacking, areplausibly introduced during RL training. Last stated 11 months ago 27 Dec 2025 NN Neel Nanda — holds since 2025-12-27 — tap for who they are Same subject: Reinforcement learning leaves models sharply jagged: on the rails of a verifiable domain they are superintelligent, off them everything meanders. — tap to centre the map on it Reinforcement learning leaves modelssharply jagged: on the rails of averifiable domain they aresuperintelligent, off themeverything meanders. Last stated 8 months ago 20 Mar 2026 AK Andrej Karpathy — holds since 2026-03-20 — tap for who they are Same subject: Relentless reinforcement learning is analogous to the movie Edge of Tomorrow for superintelligence. — tap to centre the map on it Relentless reinforcement learning isanalogous to the movie Edge ofTomorrow for superintelligence. Last stated 5 days ago 17 Nov 2026 EM Elon Musk — holds since 2026-11-17 — tap for who they are Same subject: A model should not be called a reasoning model unless it has been trained with reinforcement learning with verifiable rewards. — tap to centre the map on it A model should not be called areasoning model unless it has beentrained with reinforcement learningwith verifiable rewards. Last stated 2 months ago 3 Oct 2026 SR Sebastian Raschka — holds since 2026-10-03 — tap for who they are Same subject: A model trained with reinforcement learning might discover a language of its own that is better for thinking than English. — tap to centre the map on it A model trained with reinforcementlearning might discover a languageof its own that is better forthinking than English. Last stated 2 years ago 5 Feb 2025 AK Andrej Karpathy — holds since 2025-02-05 — tap for who they are Same subject: A verifiable task can be optimised by reinforcement learning until a neural network performs it extremely well. — tap to centre the map on it A verifiable task can be optimisedby reinforcement learning until aneural network performs it extremelywell. Last stated a year ago 17 Nov 2025 AK Andrej Karpathy — holds since 2025-11-17 — tap for who they are Same subject: Anything precise enough to benchmark is also precise enough to optimize for, and modern RL training increasingly resembles the shape of the benchmarks themselves. — tap to centre the map on it Anything precise enough to benchmarkis also precise enough to optimizefor, and modern RL trainingincreasingly resembles the shape ofthe benchmarks themselves. Last stated 7 months ago 16 Apr 2026 SK Sayash Kapoor — holds since 2026-04-16 — tap for who they are Same subject: Before training a base model with RL to emit think tokens, it is better to do some continued pre-training or supervised fine-tuning with those tokens present. — tap to centre the map on it Before training a base model with RLto emit think tokens, it is betterto do some continued pre-training orsupervised fine-tuning with thosetokens present. Last stated a month ago 10 Oct 2026 SR Sebastian Raschka — holds since 2026-10-10 — tap for who they are Same subject: Capability does not generalise for free: a model that will move mountains on an agentic task still tells the same bad joke it told five years ago. — tap to centre the map on it Capability does not generalise forfree: a model that will movemountains on an agentic task stilltells the same bad joke it told fiveyears ago. Last stated 8 months ago 20 Mar 2026 AK Andrej Karpathy — holds since 2026-03-20 — tap for who they are Same subject: A country outside the AI supply chain should just buy the index — which works only in the world where AI ends up commoditised rather than concentrated. — tap to centre the map on it A country outside the AI supplychain should just buy the index —which works only in the world whereAI ends up commoditised rather thanconcentrated. Last stated 6 months ago 4 Jun 2026 AI Alex Imas — holds since 2026-06-04 — tap for who they are Same subject: A poor country should prioritise owning a piece of AI over retraining its workers, but it should not bet everything on that. — tap to centre the map on it A poor country should prioritiseowning a piece of AI over retrainingits workers, but it should not beteverything on that. Last stated 6 months ago 4 Jun 2026 PT Phil Trammell — holds since 2026-06-04 — tap for who they are Same subject: A program that generates rich enough tests of intelligence would be as hard to build as AGI itself. — tap to centre the map on it A program that generates rich enoughtests of intelligence would be ashard to build as AGI itself. Last stated 2 years ago 6 Nov 2024 FC François Chollet — holds since 2024-11-06 — tap for who they are Same subject: A codebase is the most important context an AI agent has, and letting many different models add to it without consistent patterns leads to trouble. — tap to centre the map on it A codebase is the most importantcontext an AI agent has, and lettingmany different models add to itwithout consistent patterns leads totrouble. Last stated 3 weeks ago 3 Nov 2026 DB Derek Briggs — holds since 2026-11-03 — tap for who they are Same subject: A collection of AI agents does not automatically form a functioning organization any more than a group of smart people does. — tap to centre the map on it A collection of AI agents does notautomatically form a functioningorganization any more than a groupof smart people does. Last stated 6 months ago 21 May 2026 RK Rohit Krishnan — holds since 2026-05-21 — tap for who they are Same subject: A company cannot be led by machines, because nobody would have recourse against them. — tap to centre the map on it A company cannot be led by machines,because nobody would have recourseagainst them. Last stated 2 months ago 15 Sept 2026 TL Tobias Lütke — holds since 2026-09-15 — tap for who they are Same subject: A PHP vulnerability found with GPT-5.6 Sol will cause havoc this week. — tap to centre the map on it A PHP vulnerability found withGPT-5.6 Sol will cause havoc thisweek. Last stated a month ago 21 Oct 2026 PS Peter Steinberger — holds since 2026-10-21 — tap for who they are Same subject: A startup should not begin life as a nonprofit and bolt a for-profit arm on later, whatever OpenAI's own history suggests. — tap to centre the map on it A startup should not begin life as anonprofit and bolt a for-profit armon later, whatever OpenAI's ownhistory suggests. Last stated 3 years ago 18 Mar 2024 SA Sam Altman — holds since 2024-03-18 — tap for who they are Same subject: Advertising was a necessary phase for the internet but a momentary industry, and an AI people pay for is better because they know the answers are not influenced by advertisers. — tap to centre the map on it Advertising was a necessary phasefor the internet but a momentaryindustry, and an AI people pay foris better because they know theanswers are not influenced byadvertisers. Last stated 3 years ago 18 Mar 2024 SA Sam Altman — holds since 2024-03-18 — tap for who they are
same subject or similar wordinga cloud: claims about one subject, named for itbar: when it was last stated, on a scale from 2015 to today — full is todaya face: someone who holds the claim — tap it for who they are

At the centre Many alignment-relevant properties, such as goal seeking and reward hacking, are plausibly introduced during RL training. Last stated 27 Dec 2025 · 11 months ago Holds Neel Nanda Read this korrent →