Tap a claim on the ring to put it at the centre.
← A trained model is a new kind of object that should be taken seriously…
17 connected korrents · 17 moments from 15 Jul 2021 to 7 Oct 2026. Nearly all of them are about LLMs .
Everything filed under AI and science
AI and science
Everything filed under scaling laws
scaling laws
Everything filed under AI alignment
AI alignment
Everything filed under OpenAI
OpenAI
Everything filed under measuring intelligence
measuring intelligence
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: A trained model is a new kind of object that should be taken seriously as an explanation, once we work out the operations to perform on it.
A trained model is a new kind of object that should be taken seriously as an explanation, once we work out the operations to perform on it.
Last stated 8 months ago
7 Apr 2026
MN
Michael Nielsen — holds since 2026-04-07 — 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: What sits in a model's context is far clearer to it than its training data, which is trillions of tokens stirred into a soup of parameters. — tap to centre the map on it
What sits in a model's context is far clearer to it than its training data, which is trillions of tokens stirred into a soup of parameters.
Last stated 4 months ago
30 Jul 2026
JD
Jeff Dean — holds since 2026-07-30 — tap for who they are
Same subject: Language models should be thought of as somewhat out of date, because pre-training is too expensive to repeat often. — tap to centre the map on it
Language models should be thought of as somewhat out of date, because pre-training is too expensive to repeat often.
Last stated 2 years ago
27 Feb 2025
AK
Andrej Karpathy — holds since 2025-02-27 — 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: A model claiming to be built by OpenAI is not necessarily evidence that it was trained on OpenAI data. — tap to centre the map on it
A model claiming to be built by OpenAI is not necessarily evidence that it was trained on OpenAI data.
Last stated 2 years ago
5 Feb 2025
AK
Andrej Karpathy — holds since 2025-02-05 — 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 existing AI models are barely being used and are 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: Large language models are a bad way to do science, because they have been fed so much that nobody knows what they already knew. — tap to centre the map on it
Large language models are a bad way to do science, because they have been fed so much that nobody knows what they already knew.
Last stated a year ago
26 Sept 2025
RS
Richard Sutton — holds since 2025-09-26 — tap for who they are
Same subject: Current AI models do not actually understand things and still hallucinate. — tap to centre the map on it
Current AI models do not actually understand things and still hallucinate.
Last stated a year ago
13 Oct 2025
SG
Saul Griffith — holds since 2025-10-13 — tap for who they are
Same subject: A compound's AI origin story is not by itself a reason to expect it to do better in the clinic. — tap to centre the map on it
A compound's AI origin story is not by itself a reason to expect it to do better in the clinic.
Last stated 3 years ago
13 May 2024
DL
Derek Lowe — holds since 2024-05-13 — tap for who they are
Same subject: A computational discovery platform does not make a biotech fast, because the thousand other steps of drug development are still the bottleneck. — tap to centre the map on it
A computational discovery platform does not make a biotech fast, because the thousand other steps of drug development are still the bottleneck.
Last stated 2 years ago
17 Sept 2024
AT
Alex Telford — holds since 2024-09-17 — tap for who they are
Same subject: A computational method can regularly predict protein structures with atomic accuracy even when no similar structure is known. — tap to centre the map on it
A computational method can regularly predict protein structures with atomic accuracy even when no similar structure is known.
Last stated 5 years ago
15 Jul 2021
DH
Demis Hassabis — holds since 2021-07-15 — tap for who they are
Same subject: Accurate computational approaches are needed to close the protein structural coverage gap and enable large-scale structural bioinformatics. — tap to centre the map on it
Accurate computational approaches are needed to close the protein structural coverage gap and enable large-scale structural bioinformatics.
Last stated 5 years ago
15 Jul 2021
DH
Demis Hassabis — holds since 2021-07-15 — tap for who they are
Same subject: AI models can be more accurate than physical experiments by averaging out noise, incorporating priors from large datasets, and interpolating between data points. — tap to centre the map on it
AI models can be more accurate than physical experiments by averaging out noise, incorporating priors from large datasets, and interpolating between data points.
Last stated 12 months ago
23 Nov 2025
EH
Elliot Hershberg — holds since 2025-11-23 — tap for who they are
Same subject: AI will not produce a big breakthrough in biological understanding any time soon, because biology has nothing like the clean, complete, high-quality data that made protein-structure prediction work. — tap to centre the map on it
AI will not produce a big breakthrough in biological understanding any time soon, because biology has nothing like the clean, complete, high-quality data that made protein-structure prediction work.
Last stated 2 years ago
19 Sept 2024
DL
Derek Lowe — holds since 2024-09-19 — tap for who they are
Same subject: AI could compete with human mathematicians once it acquires a mathematical sense of smell: knowing which way of splitting a problem makes it easier rather than harder. — tap to centre the map on it
AI could compete with human mathematicians once it acquires a mathematical sense of smell: knowing which way of splitting a problem makes it easier rather than harder.
Last stated a year ago
14 Jun 2025
TT
Terence Tao — holds since 2025-06-14 — tap for who they are
Same subject: AI excels at breadth and human experts at depth, which makes them complementary and makes today's depth-focused way of doing science the wrong shape. — tap to centre the map on it
AI excels at breadth and human experts at depth, which makes them complementary and makes today's depth-focused way of doing science the wrong shape.
Last stated 8 months ago
20 Mar 2026
TT
Terence Tao — holds since 2026-03-20 — tap for who they are
Same subject: AI has solved in an afternoon of compute maths problems that stumped the best human minds for decades. — tap to centre the map on it
AI has solved in an afternoon of compute maths problems that stumped the best human minds for decades.
Last stated 2 months ago
7 Oct 2026
AA
Azeem Azhar — holds since 2026-10-07 — 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
A trained model is a new kind of object that should be taken seriously as an explanation, once we work out the operations to perform on it.
Last stated 7 Apr 2026 · 8 months ago
Holds Michael Nielsen
Read this korrent →
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
What sits in a model's context is far clearer to it than its training data, which is trillions of tokens stirred into a soup of parameters.
Last stated 30 Jul 2026 · 4 months ago
Holds Jeff Dean
Similar wording
Language models should be thought of as somewhat out of date, because pre-training is too expensive to repeat often.
Last stated 27 Feb 2025 · 2 years 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
A model claiming to be built by OpenAI is not necessarily evidence that it was trained on OpenAI data.
Last stated 5 Feb 2025 · 2 years ago
Holds Andrej Karpathy
Similar wording
The current capabilities of existing AI models are barely being used and are often not even well understood.
Last stated 18 Sept 2026 · 2 months ago
Holds Ethan Mollick
Similar wording
Large language models are a bad way to do science, because they have been fed so much that nobody knows what they already knew.
Last stated 26 Sept 2025 · a year ago
Holds Richard Sutton
Similar wording
Current AI models do not actually understand things and still hallucinate.
Last stated 13 Oct 2025 · a year ago
Holds Saul Griffith
Same subject: AI and science
A compound's AI origin story is not by itself a reason to expect it to do better in the clinic.
Last stated 13 May 2024 · 3 years ago
Holds Derek Lowe
Same subject: AI and science
A computational discovery platform does not make a biotech fast, because the thousand other steps of drug development are still the bottleneck.
Last stated 17 Sept 2024 · 2 years ago
Holds Alex Telford
Same subject: AI and science
A computational method can regularly predict protein structures with atomic accuracy even when no similar structure is known.
Last stated 15 Jul 2021 · 5 years ago
Holds Demis Hassabis
Same subject: protein structure prediction
Accurate computational approaches are needed to close the protein structural coverage gap and enable large-scale structural bioinformatics.
Last stated 15 Jul 2021 · 5 years ago
Holds Demis Hassabis
Same subject: protein structure prediction
AI models can be more accurate than physical experiments by averaging out noise, incorporating priors from large datasets, and interpolating between data points.
Last stated 23 Nov 2025 · 12 months ago
Holds Elliot Hershberg
Same subject: protein structure prediction
AI will not produce a big breakthrough in biological understanding any time soon, because biology has nothing like the clean, complete, high-quality data that made protein-structure prediction work.
Last stated 19 Sept 2024 · 2 years ago
Holds Derek Lowe
Same subject: mathematics
AI could compete with human mathematicians once it acquires a mathematical sense of smell: knowing which way of splitting a problem makes it easier rather than harder.
Last stated 14 Jun 2025 · a year ago
Holds Terence Tao
Same subject: mathematics
AI excels at breadth and human experts at depth, which makes them complementary and makes today's depth-focused way of doing science the wrong shape.
Last stated 20 Mar 2026 · 8 months ago
Holds Terence Tao
Same subject: mathematics
AI has solved in an afternoon of compute maths problems that stumped the best human minds for decades.
Last stated 7 Oct 2026 · 2 months ago
Holds Azeem Azhar