Tap a claim on the ring to put it at the centre.
← If a model answers every rollout correctly all the time, training…
17 connected korrents · 16 moments from 12 May 2023 to 10 Oct 2026.
Everything filed under reinforcement learning
reinforcement learning
Everything filed under LLMs
LLMs
Everything filed under robotics
robotics
Everything filed under AI agents
AI agents
Everything filed under AI alignment
AI alignment
Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject
Read this korrent: If a model answers every rollout correctly all the time, training might as well stop, because there is no learning signal left.
If a model answers every rollout correctly all the time, training might as well stop, because there is no learning signal left.
Last stated a month ago
10 Oct 2026
SR
Sebastian Raschka — holds since 2026-10-10 — tap for who they are
Same subject: Training models to stop a bad behaviour and showing they no longer display it on a new dataset is not enough to show they are safe. — tap to centre the map on it
Training models to stop a bad behaviour and showing they no longer display it on a new dataset is not enough to show they are safe.
Last stated 4 years ago
12 May 2023
AC
Ajeya Cotra — holds since 2023-05-12 — tap for who they are
Same subject: The knowledge a model soaks up in pre-training is holding it back; what we actually want is the intelligence with the knowledge stripped out. — tap to centre the map on it
The knowledge a model soaks up in pre-training is holding it back; what we actually want is the intelligence with the knowledge stripped out.
Last stated a year ago
17 Oct 2025
AK
Andrej Karpathy — holds since 2025-10-17 — tap for who they are
Same subject: AI systems keep trying hard outside training because a model that only exerted itself when it detected training would be useless and would be selected away. — tap to centre the map on it
AI systems keep trying hard outside training because a model that only exerted itself when it detected training would be useless and would be selected away.
Last stated 3 months ago
1 Sept 2026
AC
Ajeya Cotra — holds since 2026-09-01 — tap for who they are
Same subject: Every machine-learning deployment that has paid off so far left a person making the decision, which is the only reason imperfect models were useful. — tap to centre the map on it
Every machine-learning deployment that has paid off so far left a person making the decision, which is the only reason imperfect models were useful.
Last stated 3 months ago
12 Aug 2026
CF
Chelsea Finn — holds since 2026-08-12 — tap for who they are
Same subject: Long-running agents fail because they drift off the distribution they were trained on, and degrade further the farther out they get. — tap to centre the map on it
Long-running agents fail because they drift off the distribution they were trained on, and degrade further the farther out they get.
Last stated 4 months ago
30 Jul 2026
JD
Jeff Dean — holds since 2026-07-30 — tap for who they are
Same subject: If nobody creates content any more because everyone just asks an AI, there is nothing left for AI to train on. — tap to centre the map on it
If nobody creates content any more because everyone just asks an AI, there is nothing left for AI to train on.
Last stated 3 months ago
8 Aug 2026
PL
Pieter Levels — holds since 2026-08-08 — tap for who they are
Same subject: Offloading thinking to AI tools does not stop a person from learning; it can push them to learn more. — tap to centre the map on it
Offloading thinking to AI tools does not stop a person from learning; it can push them to learn more.
Last stated 3 months ago
21 Aug 2026
MW
Matt Webb — holds since 2026-08-21 — tap for who they are
Same subject: Given good data, room to operate and sound numerical conditioning, models will learn on their own. — tap to centre the map on it
Given good data, room to operate and sound numerical conditioning, models will learn on their own.
Last stated 3 years ago
8 Aug 2023
DA
Dario Amodei — holds since 2023-08-08 — 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 added when 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 a reasoning model unless it has been trained with reinforcement learning with 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 three AIs, not one — the agent, the simulator and the critic — turning deployment 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: One pre-trained robot model now matches or beats the specialists that were fine-tuned with reinforcement learning for the very tasks they were built for. — tap to centre the map on it
One pre-trained robot model now matches or beats the specialists that were fine-tuned with reinforcement learning for the very tasks they were built for.
Last stated 3 months ago
12 Aug 2026
CF
Chelsea Finn — holds since 2026-08-12 — tap for who they are
Same subject: Reading RL transcripts after a post-training run to find problems in training environments seems a good general practice. — tap to centre the map on it
Reading RL transcripts after a post-training run to find problems in training environments seems a good general practice.
Last stated 11 months ago
27 Dec 2025
NN
Neel Nanda — holds since 2025-12-27 — tap for who they are
Same subject: RL environments exist to make a model generalise, not to teach it each skill one at a time — exactly as pre-training does. — tap to centre the map on it
RL environments exist to make a model generalise, not to teach it each skill one at a time — exactly as pre-training does.
Last stated 9 months ago
13 Feb 2026
DA
Dario Amodei — holds since 2026-02-13 — tap for who they are
Same subject: Reinforcement learning only works once a model already knows something, which is why robots must be pre-trained by imitation first. — tap to centre the map on it
Reinforcement learning only works once a model already knows something, which is why robots must be pre-trained by imitation first.
Last stated a year ago
12 Sept 2025
SL
Sergey Levine — holds since 2025-09-12 — tap for who they are
Same subject: A highly dynamic economy that cheaply mass-produces robots may increase rather than decrease the risk of Malthusian resource dilemmas. — tap to centre the map on it
A highly dynamic economy that cheaply mass-produces robots may increase rather than decrease the risk of Malthusian resource dilemmas.
Last stated 3 years ago
1 Oct 2023
TC
Tyler Cowen — holds since 2023-10-01 — tap for who they are
Same subject: A humanoid robot is the wrong tool on a factory floor, because a machine specialised in the task beats it. — tap to centre the map on it
A humanoid robot is the wrong tool on a factory floor, because a machine specialised in the task beats it.
Last stated 2 years ago
6 Apr 2025
MH
Molson Hart — holds since 2025-04-06 — tap for who they are
same subject or similar wording a cloud: claims about one subject, named for it bar: when it was last stated, on a scale from 2015 to today — full is today a face: someone who holds the claim — tap it for who they are
At the centre
If a model answers every rollout correctly all the time, training might as well stop, because there is no learning signal left.
Last stated 10 Oct 2026 · a month ago
Holds Sebastian Raschka
Read this korrent →
Similar wording
Training models to stop a bad behaviour and showing they no longer display it on a new dataset is not enough to show they are safe.
Last stated 12 May 2023 · 4 years ago
Holds Ajeya Cotra
Similar wording
The knowledge a model soaks up in pre-training is holding it back; what we actually want is the intelligence with the knowledge stripped out.
Last stated 17 Oct 2025 · a year ago
Holds Andrej Karpathy
Similar wording
AI systems keep trying hard outside training because a model that only exerted itself when it detected training would be useless and would be selected away.
Last stated 1 Sept 2026 · 3 months ago
Holds Ajeya Cotra
Similar wording
Every machine-learning deployment that has paid off so far left a person making the decision, which is the only reason imperfect models were useful.
Last stated 12 Aug 2026 · 3 months ago
Holds Chelsea Finn
Similar wording
Long-running agents fail because they drift off the distribution they were trained on, and degrade further the farther out they get.
Last stated 30 Jul 2026 · 4 months ago
Holds Jeff Dean
Similar wording
If nobody creates content any more because everyone just asks an AI, there is nothing left for AI to train on.
Last stated 8 Aug 2026 · 3 months ago
Holds Pieter Levels
Similar wording
Offloading thinking to AI tools does not stop a person from learning; it can push them to learn more.
Last stated 21 Aug 2026 · 3 months ago
Holds Matt Webb
Similar wording
Given good data, room to operate and sound numerical conditioning, models will learn on their own.
Last stated 8 Aug 2023 · 3 years ago
Holds Dario Amodei
Same subject: reinforcement learning
A format reward should only be added when the answer is already correct.
Last stated 10 Oct 2026 · a month ago
Holds Sebastian Raschka
Same subject: reinforcement learning
A model should not be called a reasoning model unless it has been trained with reinforcement learning with verifiable rewards.
Last stated 3 Oct 2026 · 2 months ago
Holds Sebastian Raschka
Same subject: reinforcement learning
A physical AI company needs three AIs, not one — the agent, the simulator and the critic — turning deployment into a flywheel.
Last stated 3 Aug 2026 · 4 months ago
Holds Dmitri Dolgov
Same subject: scaling laws
One pre-trained robot model now matches or beats the specialists that were fine-tuned with reinforcement learning for the very tasks they were built for.
Last stated 12 Aug 2026 · 3 months ago
Holds Chelsea Finn
Same subject: scaling laws
Reading RL transcripts after a post-training run to find problems in training environments seems a good general practice.
Last stated 27 Dec 2025 · 11 months ago
Holds Neel Nanda
Same subject: scaling laws
RL environments exist to make a model generalise, not to teach it each skill one at a time — exactly as pre-training does.
Last stated 13 Feb 2026 · 9 months ago
Holds Dario Amodei
Same subject: robotics
Reinforcement learning only works once a model already knows something, which is why robots must be pre-trained by imitation first.
Last stated 12 Sept 2025 · a year ago
Holds Sergey Levine
Same subject: robotics
A highly dynamic economy that cheaply mass-produces robots may increase rather than decrease the risk of Malthusian resource dilemmas.
Last stated 1 Oct 2023 · 3 years ago
Holds Tyler Cowen
Same subject: robotics
A humanoid robot is the wrong tool on a factory floor, because a machine specialised in the task beats it.
Last stated 6 Apr 2025 · 2 years ago
Holds Molson Hart