Tap a claim on the ring to put it at the centre.
← Gradient descent finds a solution to the problems a model has seen;…
17 connected korrents · 15 moments on record from 10 Jun 2022 to 12 Aug 2026.
Everything filed under reinforcement learning
reinforcement learning
Everything filed under robotics
robotics
Everything filed under scaling laws
scaling laws
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: Gradient descent finds a solution to the problems a model has seen; nothing in the algorithm makes it pick the one that generalises well. Gradient descent finds a solution to the problems a model has seen; nothing in the algorithm makes it pick the one that generalises well. Last stated 11 months ago
26 Sept 2025
RS
Richard Sutton — holds since 2025-09-26 — tap for who they are
Same subject: You cannot build an AI that has only the capabilities you want, because the algorithms that solve the problems you want generalize to the ones you do not. — tap to centre the map on it You cannot build an AI that has only the capabilities you want, because the algorithms that solve the problems you want generalize to the ones you do not. Last stated 4 years ago
10 Jun 2022
EY
Eliezer Yudkowsky — holds since 2022-06-10 — tap for who they are
Same subject: It is the diversity of a robot data set, not its size, that produces generalization: dropping the most diverse slice hurts far more than dropping a random fifth. — tap to centre the map on it It is the diversity of a robot data set, not its size, that produces generalization: dropping the most diverse slice hurts far more than dropping a random fifth. Last stated 4 weeks ago
12 Aug 2026
CF
Chelsea Finn — holds since 2026-08-12 — tap for who they are
Same subject: A model can make an existing algorithm a hundred times faster and still cannot invent a better one, however long you give it. — tap to centre the map on it A model can make an existing algorithm a hundred times faster and still cannot invent a better one, however long you give it. Last stated 2 months ago
26 Jun 2026
NB
Noam Brown — holds since 2026-06-26 — 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 7 months ago
13 Feb 2026
DA
Dario Amodei — holds since 2026-02-13 — tap for who they are
Same subject: Models already generalise substantially from tasks that can be verified to tasks that cannot. — tap to centre the map on it Models already generalise substantially from tasks that can be verified to tasks that cannot. Last stated 7 months ago
13 Feb 2026
DA
Dario Amodei — holds since 2026-02-13 — tap for who they are
Same subject: Deep learning generalises badly, and catastrophic interference with what a network already knew is the proof of it. — tap to centre the map on it Deep learning generalises badly, and catastrophic interference with what a network already knew is the proof of it. Last stated 11 months ago
26 Sept 2025
RS
Richard Sutton — holds since 2025-09-26 — tap for who they are
Same subject: How much machine learning can help a biological problem is set by the quality of the data around that problem, not by the algorithms. — tap to centre the map on it How much machine learning can help a biological problem is set by the quality of the data around that problem, not by the algorithms. Last stated 2 years ago
19 Sept 2024
DL
Derek Lowe — holds since 2024-09-19 — 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 going well enough to look solvable, while aligning models we can no longer understand remains unsolved. Last stated 8 months ago
22 Jan 2026
JL
Jan Leike — holds since 2026-01-22 — 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 a year ago
6 Apr 2025
MH
Molson Hart — holds since 2025-04-06 — tap for who they are
Same subject: A physical AI agent has to be safe on day one, because you cannot ship something merely good enough and let users find the edge cases. — tap to centre the map on it A physical AI agent has to be safe on day one, because you cannot ship something merely good enough and let users find the edge cases. Last stated a month ago
3 Aug 2026
DD
Dmitri Dolgov — holds since 2026-08-03 — tap for who they are
Same subject: A benchmark result should be reported under a stated budget, or as a curve against test-time compute — never as a single number. — tap to centre the map on it A benchmark result should be reported under a stated budget, or as a curve against test-time compute — never as a single number. Last stated 2 months ago
26 Jun 2026
NB
Noam Brown — holds since 2026-06-26 — tap for who they are
Same subject: A lab should spend most of its compute on research rather than on building the next model, because research is where the tenfold yearly efficiency gains come from. — tap to centre the map on it A lab should spend most of its compute on research rather than on building the next model, because research is where the tenfold yearly efficiency gains come from. Last stated 6 months ago
13 Mar 2026
DP
Dylan Patel — holds since 2026-03-13 — tap for who they are
Same subject: After pre-training, post-training and test-time scaling, the fourth scaling law is agentic: multiplying AI by spawning agents, and the whole loop scales on one thing, compute. — tap to centre the map on it After pre-training, post-training and test-time scaling, the fourth scaling law is agentic: multiplying AI by spawning agents, and the whole loop scales on one thing, compute. Last stated 6 months ago
23 Mar 2026
JH
Jensen Huang — holds since 2026-03-23 — 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 optimised by reinforcement learning until a neural network performs it extremely well. Last stated 10 months ago
17 Nov 2025
AK
Andrej Karpathy — holds since 2025-11-17 — 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 for free: a model that will move mountains on an agentic task still tells the same bad joke it told five years ago. Last stated 6 months ago
20 Mar 2026
AK
Andrej Karpathy — holds since 2026-03-20 — tap for who they are
Same subject: Humans barely use reinforcement learning for intelligence — what RL they do use goes into motor tasks, not problem solving. — tap to centre the map on it Humans barely use reinforcement learning for intelligence — what RL they do use goes into motor tasks, not problem solving. Last stated 11 months ago
17 Oct 2025
AK
Andrej Karpathy — holds since 2025-10-17 — tap for who they are
At the centre
Gradient descent finds a solution to the problems a model has seen; nothing in the algorithm makes it pick the one that generalises well.
Holds RS Richard SuttonRead this korrent →
You cannot build an AI that has only the capabilities you want, because the algorithms that solve the problems you want generalize to the ones you do not.
Holds EY Eliezer Yudkowsky
It is the diversity of a robot data set, not its size, that produces generalization: dropping the most diverse slice hurts far more than dropping a random fifth.
Holds CF Chelsea Finn
A model can make an existing algorithm a hundred times faster and still cannot invent a better one, however long you give it.
Holds NB Noam Brown
RL environments exist to make a model generalise, not to teach it each skill one at a time — exactly as pre-training does.
Holds DA Dario Amodei
Models already generalise substantially from tasks that can be verified to tasks that cannot.
Holds DA Dario Amodei
Deep learning generalises badly, and catastrophic interference with what a network already knew is the proof of it.
Holds RS Richard Sutton
How much machine learning can help a biological problem is set by the quality of the data around that problem, not by the algorithms.
Holds DL Derek Lowe
Alignment of today's models is going well enough to look solvable, while aligning models we can no longer understand remains unsolved.
Holds JL Jan Leike
A highly dynamic economy that cheaply mass-produces robots may increase rather than decrease the risk of Malthusian resource dilemmas.
Holds TC Tyler Cowen
A humanoid robot is the wrong tool on a factory floor, because a machine specialised in the task beats it.
Holds MH Molson Hart
A physical AI agent has to be safe on day one, because you cannot ship something merely good enough and let users find the edge cases.
Holds DD Dmitri Dolgov
A benchmark result should be reported under a stated budget, or as a curve against test-time compute — never as a single number.
Holds NB Noam Brown
A lab should spend most of its compute on research rather than on building the next model, because research is where the tenfold yearly efficiency gains come from.
Holds DP Dylan Patel
After pre-training, post-training and test-time scaling, the fourth scaling law is agentic: multiplying AI by spawning agents, and the whole loop scales on one thing, compute.
Holds JH Jensen Huang
A verifiable task can be optimised by reinforcement learning until a neural network performs it extremely well.
Holds
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.
Holds
Humans barely use reinforcement learning for intelligence — what RL they do use goes into motor tasks, not problem solving.
Holds