Archive
Permission to compute, Mr President?
In an ongoing and public battle between Anthropic and the White House, the latter shut down the world’s most capable AI model three days after it had been released to the world. This isn't the first, nor will it be the last instance of US Government interference in the AI arms race. As Europe absorbs being locked out to the latest release, we investigate the history of US interference on technical releases and the implications on Europe’s own progress, or lack of.
Europe's cyber defences against a flood of LLM models
Anthropic built the most capable model it has ever made, then decided the unrestricted version wasn't safe to sell. The reason was cybersecurity. Which leaves a question for everyone who isn't a hyperscaler: if Europe can't compete on raw model power, can a fast-growing crop of European cyber companies build a moat out of everything that sits on top of it?
Anthropic and one of the greatest venture returns in history
Anthropic has filed to go public at close to a trillion dollars. For the investors who got in early, the listing will unlock one of the largest fortunes in venture history. This edition is about how that fortune was built, why almost none of it was reachable by the public, and why the investor who buys at the listing inherits a different and far less forgiving set of arithmetic.
The battle for AI in professional services
OpenAI, Anthropic, Harvey, Legora and a wave of vertical specialists are now openly fighting for the same prize: the multi-trillion-dollar spend that flows through legal, banking, accounting and consulting every year. Here is who is doing what, and where the trade is going.
