Tap a claim on the ring to put it at the centre.
← There is no real impediment to models running their own research loop…
17 connected korrents · 13 moments on record from 12 Feb 2010 to 3 Sept 2026.
Everything filed under reinforcement learning
reinforcement learning
Everything filed under scaling laws
scaling laws
Everything filed under taste
taste
Everything filed under design
design 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: There is no real impediment to models running their own research loop and improving themselves at a far more rapid rate. There is no real impediment to models running their own research loop and improving themselves at a far more rapid rate. Last stated a month ago
30 Jul 2026
JD
Jeff Dean — holds since 2026-07-30 — 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: Models have no research taste yet, which is why they complement researchers rather than replace them. — tap to centre the map on it Models have no research taste yet, which is why they complement researchers rather than replace them. Last stated 2 months ago
26 Jun 2026
NB
Noam Brown — holds since 2026-06-26 — tap for who they are
Same subject: Learned approximations of slow simulators change what science is possible, by turning a six-month screening run into something you do over lunch. — tap to centre the map on it Learned approximations of slow simulators change what science is possible, by turning a six-month screening run into something you do over lunch. Last stated a month ago
30 Jul 2026
JD
Jeff Dean — holds since 2026-07-30 — tap for who they are
Same subject: Models look far better on evals than they are in the world because researchers, inadvertently, take inspiration from the evals when they build RL environments. — tap to centre the map on it Models look far better on evals than they are in the world because researchers, inadvertently, take inspiration from the evals when they build RL environments. Last stated 9 months ago
25 Nov 2025
IS
Ilya Sutskever — holds since 2025-11-25 — tap for who they are
Same subject: Coding models are worst at exactly the thing an AI research explosion would need: code that has never been written before. — tap to centre the map on it Coding models are worst at exactly the thing an AI research explosion would need: code that has never been written before. Last stated 11 months ago
17 Oct 2025
AK
Andrej Karpathy — holds since 2025-10-17 — tap for who they are
Same subject: A narrow, highly accurate model can beat a general one inside its own domain, as AlphaFold did, and materials science and chip design are next. — tap to centre the map on it A narrow, highly accurate model can beat a general one inside its own domain, as AlphaFold did, and materials science and chip design are next. Last stated a month ago
30 Jul 2026
JD
Jeff Dean — holds since 2026-07-30 — tap for who they are
Same subject: The least verifiable part of AI research is the judgement call about what goes into the one big training run. — tap to centre the map on it The least verifiable part of AI research is the judgement call about what goes into the one big training run. Last stated 4 weeks ago
11 Aug 2026
RG
Ryan Greenblatt — holds since 2026-08-11 — 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: 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
Same subject: A first draft should not be graded good or bad; it is only the material that taste then gets to act on. — tap to centre the map on it A first draft should not be graded good or bad; it is only the material that taste then gets to act on. Last stated a month ago
6 Aug 2026
GS
George Saunders — holds since 2026-08-06 — tap for who they are
Same subject: A lot of people can match a framework for taste; almost nobody can create one, and creating one is the rare skill. — tap to centre the map on it A lot of people can match a framework for taste; almost nobody can create one, and creating one is the rare skill. Last stated 11 months ago
16 Oct 2025
DF
Dylan Field — holds since 2025-10-16 — tap for who they are
Same subject: A product needs a soul, and for that it needs one person of great taste who is its living, breathing aspect and gets furious about every small detail. — tap to centre the map on it A product needs a soul, and for that it needs one person of great taste who is its living, breathing aspect and gets furious about every small detail. Last stated 17 years ago
12 Feb 2010
KS
Karri Saarinen — holds since 2010-02-12 — 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: 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: AI model capability progress is not going to slow down soon — tap to centre the map on it AI model capability progress is not going to slow down soon Last stated 4 days ago
3 Sept 2026
AR
Armin Ronacher — holds since 2026-09-03 — tap for who they are
At the centre
There is no real impediment to models running their own research loop and improving themselves at a far more rapid rate.
Holds JD Jeff DeanRead this korrent →
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
Models have no research taste yet, which is why they complement researchers rather than replace them.
Holds NB Noam Brown
Learned approximations of slow simulators change what science is possible, by turning a six-month screening run into something you do over lunch.
Holds JD Jeff Dean
Models look far better on evals than they are in the world because researchers, inadvertently, take inspiration from the evals when they build RL environments.
Holds IS Ilya Sutskever
Coding models are worst at exactly the thing an AI research explosion would need: code that has never been written before.
Holds
A narrow, highly accurate model can beat a general one inside its own domain, as AlphaFold did, and materials science and chip design are next.
Holds JD Jeff Dean
The least verifiable part of AI research is the judgement call about what goes into the one big training run.
Holds RG Ryan Greenblatt
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
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
A first draft should not be graded good or bad; it is only the material that taste then gets to act on.
Holds GS George Saunders
A lot of people can match a framework for taste; almost nobody can create one, and creating one is the rare skill.
Holds DF Dylan Field
A product needs a soul, and for that it needs one person of great taste who is its living, breathing aspect and gets furious about every small detail.
Holds KS Karri Saarinen
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
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
AI model capability progress is not going to slow down soon
Holds