Token for Token

'AI' just means LLMs now

There used to be many possible candidates for how humans might build intelligent systems. Now there's only one.

Aug 2 • 

Jack Morris

24

4

July 2025

All AI Models Might Be The Same

what can language model embeddings tell us about whales speech, and decoding ancient texts? (on The Platonic Representation Hypothesis and the idea of…

Jul 17 • 

Jack Morris

126

9

How to scale RL to 10^26 FLOPs

A roadmap for RL-ing LLMs on the entire Internet

Jul 10 • 

Jack Morris

53

11

June 2025

Superintelligence, from First Principles

There are only two ways to learn: supervised learning and reinforcement learning. Which one will give us superintelligence?

Jun 18 • 

Jack Morris

35

4

The Case for More Ambition

why AI researchers should dream bigger and publish less

Jun 9 • 

Jack Morris

51

9

April 2025

There Are No New Ideas in AI… Only New Datasets

LLMs were invented in four major developments... all of which were datasets

Apr 9 • 

Jack Morris

279

25

February 2025

CoPilot for Everything

Our employers have all the data they need to train AI models to replace us. How long will it be until this actually happens?

Feb 28 • 

Jack Morris

20

2

Please Stop Talking About AGI

Why I think Yann Lecun was right about LLMs (but perhaps only by accident)

Feb 21 • 

Jack Morris

54

6