I had gone to Hugging Face for something else entirely. I ended up spending the evening reading the report of the first cyber-intrusion carried out, from start to finish, by an autonomous artificial intelligence. This is the story of that intrusion – but to tell it properly you first have to know what the platform that was hit actually is, how “open” AI models changed the landscape, what autonomous agents are, and why the alignment problem, which seemed like a thing for philosophers, has just become a matter for the incident-response handbook. If you're in a hurry, you can skip straight to the anatomy of the intrusion. But if there's one thing I'd ask you to read to the end, it's the twist: because five days after this case was published, the author of the attack confessed – and it's not who any of us would have bet on.

A few days ago Anthropic released Claude Fable 5 and its older sibling Mythos 5. Frontier, agentic models, able to reason for hours over enormous codebases, to use tools autonomously, to behave almost like a senior software engineer. Fable 5 came out on Tuesday 9 June; by Friday the 12th, after about 72 hours of life, it was already gone. For a few hours – actually, for a few days – it was available to everyone. Then came the silence.

Leo XIV's encyclical Magnifica Humanitas came out on 15 May, and within a week I had already read more or less every possible word of praise. The Catholics of the left (read: catto-communists) celebrated it for its explicit anti-capitalism; the critics of technology (read: techno-sceptics) for its warning against Big Tech; the mainstream press for the pop citations – Tolkien, Beethoven, Schindler's List; even a few self-declared atheists, scattered across social media, tipped their hats at the lucidity with which a Pope names the concentration of computational power in the hands of a few. At the presentation, in the Synod Hall, Chris Olah sat among the speakers – co-founder of Anthropic and head of research on AI interpretability. This is not a detail: it is the signature on a document that wants to be taken seriously even by those who actually build the models. Of praise, in short, I have read enough. I, however, want to do the opposite exercise.

3:00 AM. Another one of those nights where my brain decided sleep was overrated. After my usual nocturnal walk through the streets of a remote Scottish town—where even a fox observed me with that “humans are weird” look—I sat back down at my server. Just a quick scan of my RSS feeds, I told myself, then I can start work. When...

We backed up Spotify (metadata and music files). It's distributed in bulk torrents (~300TB), grouped by popularity. This release includes the largest publicly available music metadata database with 256 million tracks and 186 million unique ISRCs. It's the world's first “preservation archive” for music which is fully open (meaning it can easily be mirrored by anyone with enough disk space), with 86 million music files, representing around 99.6% of listens.

The news came from Anna's Archive—the world's largest pirate library—which had just scraped Spotify's entire catalog. Not just metadata, but also the audio files. 86 million tracks, 300 terabytes. I stopped to reread those numbers, then thought: holy shit, how big is this thing?

When the world woke up astonished in November 2022 to this “magical” chatbot, few realized that this magic was the result of decades of research. The history of artificial intelligence begins in 1943, when Warren McCulloch and Walter Pitts proposed the first mathematical model of an artificial neuron. In 1956, at the Dartmouth Conference, John McCarthy coined the term “Artificial Intelligence” and the discipline was officially born.

The '60s and '70s were characterized by excessive optimism: people thought strong AI was just around the corner. Two “AI winters” followed – periods when funding disappeared and research slowed – because promises weren't materializing. But some continued working in the shadows. Geoffrey Hinton, Yann LeCun, Yoshua Bengio – those we now call the “godfathers of deep learning” – continued their studies on neural networks when no one believed in them anymore.