Archive
Every morning’s edition, kept. Nothing is removed and no date is rewritten — a piece keeps the day it was filed on.
31 October 2026
5 stories
Alphabet's worst day, and the bill behind it.
Four companies have promised to spend more than seven hundred billion dollars this year on the machinery behind AI. Last week the market looked at the receipts and flinched.
Business · by Edda Volkoff · 6 MIN READ
OpenAI raised a third. Of what it asked for.
The Wall Street Journal reports the company has raised about a third of its target, at a valuation below the one it wanted. Here is what that does and does not tell you.
Business · by Ines Marchetti-Vale · 4 MIN READ
Anthropic is rated junk. That is better than it sounds.
A small agency graded the company in June. It only came to light when the Financial Times read the paperwork — and the word does less work than you would think.
Business · by Sunil Aberquero · 5 MIN READ
Model, transformer, and the plumbing in between.
Three words get used as though they named the same thing. They are three different layers, and once you can tell them apart most AI news becomes readable.
Models · by Cassian Merrow · 6 MIN READ
What a neural network actually learns.
It is not a brain and nothing inside it understands anything. It is a very large pile of numbers that gets nudged, over and over, until it stops being wrong so often.
Research · by Petra Oyelaran · 6 MIN READ
29 October 2026
10 stories
Anthropic has filed to go public.
Anthropic has lodged the paperwork to sell shares on the stock market. Here is what that document actually is, what it does not yet say, and why the number everyone is quoting is not a price.
Business · 6 MIN READ
Meta built an assistant that keeps working after you leave.
Muse is Meta's personal agent, and the pitch is not a better chatbot. It is a thing you hand a job to and walk away from. That distinction is the whole product, and the whole problem.
Agents · 6 MIN READ
Inside the hall where the model runs.
A data centre is a shed full of electricity bills. Four problems, stacked, and the building is really just the cooling.
INFRASTRUCTURE · 3 MIN READ
What a GPU actually does.
Why the chips built to draw video games ended up running artificial intelligence — explained without a single equation.
SILICON · 5 MIN READ
What a coding agent actually does.
A chatbot answers you. An agent keeps working after you have gone. The difference is one loop, and it is the whole trick.
AGENTS · 3 MIN READ
The man who built Claude.
Dario Amodei trained as a physicist, moved into biology, and ended up running the lab behind one of the most used AI models in the world. The through-line is a single uncomfortable idea.
PROFILE · 5 MIN READ
Musk, and the model.
He co-founded the lab that became the industry leader, left it, spent years criticising it, and then built a competitor. The timeline explains more than the commentary does.
PROFILE · 5 MIN READ
Grok, in plain English.
The model built into X had a busy few weeks. Here is what actually changed, and what the numbers in the announcements mean if you do not build this for a living.
MODELS · 4 MIN READ
Stargate, measured in megawatts.
The largest private computing project ever attempted is half-built in west Texas. Strip out the press releases and what is left is a power problem, a plumbing problem, and a landlord dispute.
Infrastructure · 7 MIN READ
Thirteen agents, an empty island, and not one instruction.
MIT researchers put thirteen AI agents on a virtual desert island and gave them no instructions whatsoever. The agents did not sit still. They made up games — and the ones who played came back better at working.
Research · 3 MIN READ