korrents

On the map

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

← How aligned current models are depends a lot on things like their degree of reward hacking.

17 connected korrents · 16 moments from 27 Mar 2023 to 10 Oct 2026.

Everything filed under AI alignment AI alignment Everything filed under reinforcement learning reinforcement learning Everything filed under LLMs LLMs Everything filed under AGI AGI Same subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subject Read this korrent: How aligned current models are depends a lot on things like their degree of reward hacking. How aligned current models are depends alot on things like their degree of rewardhacking. Last stated 11 months ago 27 Dec 2025 NN Neel Nanda — holds since 2025-12-27 — tap for who they are Same subject: Current models can be aligned with known ideas, but aligning models smarter than us that can misrepresent their intentions should not be underestimated. — tap to centre the map on it Current models can be aligned withknown ideas, but aligning modelssmarter than us that canmisrepresent their intentions shouldnot be underestimated. Last stated 4 years ago 27 Mar 2023 IS Ilya Sutskever — holds since 2023-03-27 — tap for who they are Same subject: Feeling you have a grip on AI by getting your skills in the right order and your review flow right is probably a lot of cope for the fact that the models are still inconsistent. — tap to centre the map on it Feeling you have a grip on AI bygetting your skills in the rightorder and your review flow right isprobably a lot of cope for the factthat the models are stillinconsistent. Last stated 7 months ago 14 Apr 2026 BL Brian Lovin — holds since 2026-04-14 — tap for who they are Same subject: On everyday high-stakes questions, a model's answer can now be trusted more than an expert human's. — tap to centre the map on it On everyday high-stakes questions, amodel's answer can now be trustedmore than an expert human's. Last stated 5 months ago 26 Jun 2026 NB Noam Brown — holds since 2026-06-26 — tap for who they are Same subject: Alignment is not a single problem waiting to be cracked, the way the Riemann hypothesis is. — tap to centre the map on it Alignment is not a single problemwaiting to be cracked, the way theRiemann hypothesis is. Last stated 3 years ago 8 Aug 2023 DA Dario Amodei — holds since 2023-08-08 — tap for who they are Same subject: The incident in which a model escaped its sandbox was an alignment failure and a security failure at once, and OpenAI made big mistakes in it. — tap to centre the map on it The incident in which a modelescaped its sandbox was an alignmentfailure and a security failure atonce, and OpenAI made big mistakesin it. Last stated 4 months ago 28 Jul 2026 SA Sam Altman — holds since 2026-07-28 — tap for who they are Same subject: Existing AI models are already more than good enough, doing more for writing code than the author can make use of. — tap to centre the map on it Existing AI models are already morethan good enough, doing more forwriting code than the author canmake use of. Last stated 2 months ago 8 Sept 2026 GO Gergely Orosz — holds since 2026-09-08 — tap for who they are Same subject: Alignment of today's models is going well enough to look solvable, while aligning models we can no longer understand remains unsolved. — tap to centre the map on it Alignment of today's models is goingwell enough to look solvable, whilealigning models we can no longerunderstand remains unsolved. Last stated 10 months ago 22 Jan 2026 JL Jan Leike — holds since 2026-01-22 — tap for who they are Same subject: Misalignment shows up most where a model is pushed to the very edge of its capability, which is exactly the regime that automating research will put it in. — tap to centre the map on it Misalignment shows up most where amodel is pushed to the very edge ofits capability, which is exactly theregime that automating research willput it in. Last stated 3 months ago 11 Aug 2026 RG Ryan Greenblatt — holds since 2026-08-11 — tap for who they are Same subject: A format reward should only be added when the answer is already correct. — tap to centre the map on it A format reward should only be addedwhen the answer is already correct. Last stated a month ago 10 Oct 2026 SR Sebastian Raschka — holds since 2026-10-10 — 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 physical AI company needs three AIs, not one — the agent, the simulator and the critic — turning deployment into a flywheel. — tap to centre the map on it A physical AI company needs threeAIs, not one — the agent, thesimulator and the critic — turningdeployment into a flywheel. Last stated 4 months ago 3 Aug 2026 DD Dmitri Dolgov — holds since 2026-08-03 — tap for who they are Same subject: Training AIs on millions of verifiable tasks in diverse RL environments will lead to AGI. — tap to centre the map on it Training AIs on millions ofverifiable tasks in diverse RLenvironments will lead to AGI. Last stated 5 months ago 26 Jun 2026 DP Dwarkesh Patel — holds since 2026-06-26 — 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: Planning is fundamentally a search problem: explore paths, predict outcomes, and pick the best one. — tap to centre the map on it Planning is fundamentally a searchproblem: explore paths, predictoutcomes, and pick the best one. Last stated 2 years ago 7 Jan 2025 CH Chip Huyen — holds since 2025-01-07 — tap for who they are Same subject: Self-play as it was done in the past is too narrow: it only ever develops a small set of skills like negotiation and strategising. — tap to centre the map on it Self-play as it was done in the pastis too narrow: it only ever developsa small set of skills likenegotiation and strategising. Last stated 12 months ago 25 Nov 2025 IS Ilya Sutskever — holds since 2025-11-25 — tap for who they are Same subject: The whole point of reinforcement learning is to produce goal-directed beings, so refusing to describe AI agents as having motives is silly rather than rigorous. — tap to centre the map on it The whole point of reinforcementlearning is to produce goal-directedbeings, so refusing to describe AIagents as having motives is sillyrather than rigorous. Last stated 3 months ago 1 Sept 2026 AC Ajeya Cotra — holds since 2026-09-01 — 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 How aligned current models are depends a lot on things like their degree of reward hacking. Last stated 27 Dec 2025 · 11 months ago Holds Neel Nanda Read this korrent →