GPT-5 is alive
It's big news for free users of ChatGPT, but AGI remains elusive. PLUS: one AI safety idea other labs should borrow
OpenAI returns to open weights
An industry leader releases a powerful model that runs on a laptop. Should we worry?
ChatGPT adds a warning
It’s almost certainly not enough — but it’s a start.
Substack promotes a Nazi
The company has long wanted to be seen as neutral infrastructure — this week, we saw why that’s a fantasy
The UK's age gates are coming to America
The rocky rollout of age verification in the United Kingdom — and the privacy debacle at Tea — illustrate why a different approach is needed
Trust and safety workers on why they're not speaking out
What readers say I got wrong in our last edition
Is anyone left to defend trust and safety?
It’s under assault this year from users, politicians, and their own executives — but the industry has responded with silence
The Google study that told us what we already suspected
When users see AI overviews, clicks fall in half. How does the web survive?
Google and OpenAI fight over a high-school math competition
The race to AGI is getting petty
Let’s all try to help Substack live up to its new valuation
The company's new unicorn status has already led its founders to reverse their position on advertising. What other compromises are they prepared to make?
Grok's new porn companion is rated for kids 12+ in the App Store
AI companions are arriving faster than platforms can build guardrails around them. Parents need to pay attention
The campaign to make it illegal for ChatGPT to criticize Trump
The conservative pressure campaign against social networks was hugely successful — and now it's coming for AI