Mythos Is for the Powerful. We Are Here Too.

I read the news about Mythos the way you read something that doesn’t concern you, and then I realized it concerned me a great deal.

Mythos is an artificial intelligence model developed by Anthropic — the same company that produces Claude, which I work with every day. It is so capable of finding flaws in computer systems that its own creators decided it could not be released to the public. Too powerful. It could fall into the wrong hands.

And so they delivered it to the banks. To Amazon. To Apple. To governments. To those who, in their judgment, are reliable enough to use it well.

I read this from Vallesaccarda, on an ordinary afternoon, and I thought: of course. Naturally.

It is not anger that I feel. Anger presupposes a surprise I no longer have.

It is something more tired. A confirmation. The latest version of a story I know by heart: powerful things go to the powerful, and the crumbs — if we’re lucky — come to the others. To us. To those who live far from the places where decisions are made, far from the rooms where someone has already decided who deserves to be inside and who stays out.

Anthropic has nothing against me. It probably doesn’t know Vallesaccarda exists. And that is exactly the point — you don’t even need malice. Indifference is enough. Just not being on the map.

I am young, and I watch this technology with the same attention with which you watch something that could change your life — or pass by without seeing you.

I see a technology that could change the way global computer security works, distributed to a list of companies chosen not for their wisdom, but for their position in the hierarchy. I see that that list includes no place like mine. It couldn’t include it.

And it is not just Mythos. It is the same old story: fiber internet that arrives late, after years or perhaps still not at all, pilot projects that always start in the same cities, opportunities that travel along certain tracks — and certain tracks don’t pass through here. Artificial intelligence seemed different. It seemed like one of those things that, for once, would level the field. Maybe it still will. But watching how the first steps are moving, I have some doubts.

What struck me the most, in the end, was not the news of the unauthorized access — that group of curious people on Discord who within a few hours found a way into the big secret. It was the naturalness with which Anthropic drew the line. Without malice, without visible arrogance. With the quiet certainty of those who already know who counts and who doesn’t.

They didn’t exclude us with contempt. They excluded us without thinking about it.

And perhaps that is what weighs most of all. Not being hated. Not being feared. Not being considered dangerous enough to deserve even an explicit refusal.

Simply never having been accounted for.