Transformer

Home
Weekly Briefing
Analysis
About

Analysis

Decentralized training isn't a policy nightmare — yet
Governments should still be able to keep track of who's training frontier models — though a shift to reinforcement learning could make that harder
 • 
Lynette Bye
The embarrassing failure of the Paris AI Summit
Experts are sounding the alarm — but governments simply won’t listen
 • 
Shakeel Hashim
Leaked: this is the AI Action Summit statement
The statement, set to be signed by countries next week, is a 'wasted opportunity', experts say — and looks unlikely to be signed by US officials
 • 
Shakeel Hashim
Exclusive: Americans overwhelmingly support AI safety mandates, new poll finds
The public would rather ban AI development than have no regulation at all
 • 
Shakeel Hashim
The way we evaluate AI model safety might be about to break
As systems become more capable, researchers think we need a new type of safety evaluation
 • 
Lynette Bye
Does Elon still care about AI safety?
He’s doing a terrible job of showing it.
 • 
Shakeel Hashim
New AI export controls have leaked. Here’s what you need to know.
The new rules are squarely focused on the frontier
 • 
Shakeel Hashim
The media needs to start taking AGI seriously
An essay for Nieman Lab's 2025 Predictions series
 • 
Shakeel Hashim
OpenAI's new model tried to avoid being shut down
o1 attempted to exfiltrate its weights to avoid being shut down
 • 
Shakeel Hashim
The 2024 Transformer Gift Guide
The must-buy items for the AI policy people in your life
 • 
Shakeel Hashim
Synthetic data is more useful than you think
Fears of ‘model collapse’ are overstated — at least for now
 • 
Lynette Bye
New OpenAI emails reveal a long history of mistrust
Greg Brockman and Ilya Sutskever had questions about Sam Altman's intentions as early as 2017
 • 
Shakeel Hashim
© 2025 Shakeel Hashim
Privacy ∙ Terms ∙ Collection notice
Start WritingGet the app
Substack is the home for great culture