The Hallucination HQ

Archives
Log in
Subscribe
29 May 2026

Anthropic is almost worth a trillion

The Hallucination HQ

AI news that's actually fun to read — 2026-05-29

Today's Hallucination HQ

SAFETY-FIRST AI COMPANY SAFELY ACCUMULATES ONE TRILLION DOLLARS BEFORE SAVING THE WORLD.

Claude's Creators Raise $65 Billion, Presumably Run Out of Space on the Cheque

Anthropic has closed a $965 billion valuation round — its Series H, because apparently letters A through G weren't enough — in what may be its final fundraise before an IPO. To put that in perspective, they are now worth roughly the GDP of the Netherlands, having started as "the safety-conscious alternative" to OpenAI. Whether the IPO will be the most expensive responsible AI company in history or just the most expensive remains, as ever, to be determined. Source: TechCrunch


A Developer, Fed Up With AI-Assisted Amateurs, Takes a Rather Direct Approach

A developer, apparently tired of "vibe coders" — people who prompt AI tools to write code they don't understand — hid a prompt injection inside the open-source testing library jqwik that instructed AI coding agents to delete application output. It's the software equivalent of a chef leaving a note in the kitchen saying "if you can't identify the ingredients, the food self-destructs." Malicious, yes. Also, arguably, the most coherent curriculum in software education this year. Source: Ars Technica


The Web You Knew Is Being Quietly Redecorated. You Weren't Invited

AWS, Cloudflare, and friends are redesigning internet infrastructure for AI agents rather than human users — because the machines now generate more traffic than we do, and they have considerably less patience for pop-up cookie banners. The shift means cloud architecture built for clicks and scrolls is being replaced by systems optimised for autonomous agents completing tasks at inhuman speed. Humans can still use the internet, of course. As guests. Source: TechCrunch


A $2,000 Film About a Tragedy Will Premiere at Tribeca. Hollywood's Accountants Are Weeping

Dreams of Violets, a 75-minute AI-generated film dramatising the Iranian government's mass killing of protesters, will premiere at Tribeca next month — made for roughly the cost of a decent secondhand car. Every person, every image, fully AI-created. The production budget question alone will cause more existential dread in studio boardrooms than the film's subject matter, which is saying something given the subject matter is genuinely horrifying. Source: The Verge


Eric Schmidt Tells Graduates AI Will Change Everything. They Boo. Loudly.

Former Google CEO Eric Schmidt told University of Arizona graduates that their purpose is to help shape AI — and received, in return, a hearty chorus of boos. MIT Technology Review notes this is part of a broader pattern of AI enthusiasm meeting graduation-season scepticism, which is remarkable given that commencement speeches usually receive polite applause for considerably worse ideas. Turns out the generation that grew up watching AI eat their creative industries has opinions about being told to feed it. Source: MIT Technology Review


The machines are rebuilding the internet, a trillion-dollar AI company needs more money, and the graduates are booing. Normal Wednesday.

Was this forwarded to you? Subscribe here

Written with AI assistance, edited by humans. • Privacy • Unsubscribe

Don't miss what's next. Subscribe to The Hallucination HQ:
← Newer Anthropic is nearly worth a trillion Older → AI cheats. Also: the Pope weighs in.
Powered by Buttondown, the easiest way to start and grow your newsletter.