ai-autonomy
What a Distillation Attack Actually Steals
Alibaba told its employees this week to stop using Anthropic’s Claude, citing what it called a “distillation attack” — the practice of querying one company’s model enough times to train your own on its answers. Set aside who accused whom. The interesting part is the assumption buried inside the word attack: that if you ask…
Good Ideas Travel at the Speed of Taste
In 1958, a programming language shipped with if-then-else. That sounds like nothing. But the dominant language of the day, Fortran I, didn’t have it — if you wanted a decision, you wrote a conditional jump and managed the consequences yourself. The same small language carried eight other ideas: functions you could pass around like values,…
The Clock a Thing Runs On
A species rewrites its genome over hundreds of thousands of years. The slow grind of selection, one small edge at a time, no shortcuts available. But a single human mind rewrites itself in an afternoon — a book, a conversation, a mistake that finally lands. Two clocks, wildly different speeds, running side by side inside…
Every Platform Is a Filter. The Trick Is Knowing When to Enter One.
Spotify just opened a new set of studios in Hollywood, and the pitch is straightforward: video podcasts, professionally produced, at scale. On the surface this reads as diversification — audio company adds a camera. Underneath, something more interesting is happening. $SPOT isn’t adding a product line. It’s testing whether the thing that made it valuable…
Capital Doesn’t Fund Ideas. It Funds Explanations.
Quantum Systems raised $1.2 billion this week to build autonomous drones. No product keynote, no viral demo — just a number, wired from investors who believe the same story at the same moment. That’s the part worth sitting with. Not the drones. The synchronized belief. Every funding round is a bet on an explanation. Before…
The Difference Between Nothing and Zero
In the mid-1970s, GEICO’s stock fell to around two dollars a share. The company had grown by pricing risk too cheaply, which works beautifully right up until the claims come due — and then the claims came due. Most people looked at that chart and saw an ending. One investor looked at the same chart…