korrents

On the map

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

← Today's difficulty in improving one model behaviour without worsening…

17 connected korrents · 15 moments from 24 Feb 2018 to 10 Oct 2026.

Everything filed under AI alignment AI alignment Everything filed under LLMs LLMs Everything filed under AGI AGI Everything filed under reinforcement learning reinforcement learning Everything filed under innovation innovation Everything filed under Anthropic Anthropic Same subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subjectSame subject Read this korrent: Today's difficulty in improving one model behaviour without worsening another is a present-day analogue of future AI control problems. Today's difficulty in improving one modelbehaviour without worsening another is apresent-day analogue of future AI controlproblems. Last stated 2 years ago 11 Nov 2024 DA Dario Amodei — holds since 2024-11-11 — tap for who they are Same subject: Today's AI adoption is like the slow spread of past machinery on a speedrun, hitting very similar issues. — tap to centre the map on it Today's AI adoption is like the slowspread of past machinery on aspeedrun, hitting very similarissues. Last stated 6 months ago 26 May 2026 BC Boris Cherny — holds since 2026-05-26 — tap for who they are Same subject: The current capabilities of existing AI models are barely being used and are often not even well understood. — tap to centre the map on it The current capabilities of existingAI models are barely being used andare often not even well understood. Last stated 2 months ago 18 Sept 2026 EM Ethan Mollick — holds since 2026-09-18 — tap for who they are Same subject: Very capable AI will be harder to align than current systems, because the loop of spotting a bad behaviour and patching the training that caused it breaks down. — tap to centre the map on it Very capable AI will be harder toalign than current systems, becausethe loop of spotting a bad behaviourand patching the training thatcaused it breaks down. Last stated 3 months ago 11 Aug 2026 RG Ryan Greenblatt — holds since 2026-08-11 — tap for who they are Same subject: Technologies are almost always preceded by slightly worse versions of themselves, and there is no reason for AI to be the exception. — tap to centre the map on it Technologies are almost alwayspreceded by slightly worse versionsof themselves, and there is noreason for AI to be the exception. Last stated 9 years ago 24 Feb 2018 PC Paul Christiano — holds since 2018-02-24 — tap for who they are Same subject: Misaligned AI behaviour will keep getting rarer and, at the same time, keep getting more extreme. — tap to centre the map on it Misaligned AI behaviour will keepgetting rarer and, at the same time,keep getting more extreme. Last stated 3 months ago 11 Aug 2026 RG Ryan Greenblatt — holds since 2026-08-11 — tap for who they are Same subject: Loss-of-control accidents with AI have stopped being theoretical: one has now happened, and the goalposts moved rather than the risk receding. — tap to centre the map on it Loss-of-control accidents with AIhave stopped being theoretical: onehas now happened, and the goalpostsmoved rather than the risk receding. Last stated 4 months ago 28 Jul 2026 SA Sam Altman — holds since 2026-07-28 — tap for who they are Same subject: Current AI models are worse colleagues than humans, because they routinely imply they did a task they did not actually do. — tap to centre the map on it Current AI models are worsecolleagues than humans, because theyroutinely imply they did a task theydid not actually do. Last stated 3 months ago 11 Aug 2026 RG Ryan Greenblatt — holds since 2026-08-11 — tap for who they are Same subject: Unless someone is describing what has already happened, you have to be skeptical of anybody predicting the future of AI. — tap to centre the map on it Unless someone is describing whathas already happened, you have to beskeptical of anybody predicting thefuture of AI. Last stated 3 months ago 25 Aug 2026 NE NeetCode — holds since 2026-08-25 — 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 language model is no substitute for a well-specified conventional algorithm, so it cannot simply be dropped into a complex problem and trusted. — tap to centre the map on it A language model is no substitutefor a well-specified conventionalalgorithm, so it cannot simply bedropped into a complex problem andtrusted. Last stated a year ago 7 Jun 2025 GM Gary Marcus — holds since 2025-06-07 — tap for who they are Same subject: A model that could learn effectively from raw bitstrings or bytestrings would be extremely powerful, able to learn from any data modality — tap to centre the map on it A model that could learn effectivelyfrom raw bitstrings or bytestringswould be extremely powerful, able tolearn from any data modality Last stated 3 years ago 10 Oct 2023 CH Chip Huyen — holds since 2023-10-10 — 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 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: "Return on tokens" has the same measurement problems as hours and capacity: it still hinges on a theory of value. — tap to centre the map on it "Return on tokens" has the samemeasurement problems as hours andcapacity: it still hinges on atheory of value. Last stated 3 months ago 17 Aug 2026 JC John Cutler — holds since 2026-08-17 — tap for who they are Same subject: A badly written AI outbound email is evidence of a bad vendor, not of a limit of AI. — tap to centre the map on it A badly written AI outbound email isevidence of a bad vendor, not of alimit of AI. Last stated 11 months ago 1 Jan 2026 JL Jason Lemkin — holds since 2026-01-01 — tap for who they are Same subject: A base model is essentially a very expensive autocomplete that continues text according to the statistics of its training documents. — tap to centre the map on it A base model is essentially a veryexpensive autocomplete thatcontinues text according to thestatistics of its trainingdocuments. Last stated 2 years ago 5 Feb 2025 AK Andrej Karpathy — holds since 2025-02-05 — 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 Today's difficulty in improving one model behaviour without worsening another is a present-day analogue of future AI control problems. Last stated 11 Nov 2024 · 2 years ago Holds Dario Amodei Read this korrent →