Journal
Essays on the road to superintelligence.
Definitions first, then the argument for speed, the inputs behind progress, agents, the engineering of safety, and the problem of measurement.
What we mean by superintelligence
Three different things hide behind one word: thinking faster, thinking together, and thinking better. Only the third is new.
2 min → 5 October 2026AGI and superintelligence are two different claims
One says a machine can match us. The other says it can leave us behind. They need different evidence, and they are often confused.
2 min → 5 October 2026The intelligence explosion, six decades on
In 1965 a statistician described a machine that improves itself. The argument has barely changed since. The evidence has.
2 min → 5 October 2026Compute, data, algorithms: the three inputs
Every jump in AI capability so far has come from more of one of three things. Knowing which one is scarce tells you where progress will stall.
2 min → 5 October 2026What agents change
A model that answers is a tool. A model that acts is a colleague you have to supervise. The difference is one loop, and it changes the arithmetic of reliability.
2 min → 5 October 2026Alignment is an engineering problem first
Before the philosophy there are four practical jobs: say what you want, make it hold under pressure, stay able to intervene, and look inside.
2 min → 5 October 2026How do you measure a mind?
The SI has seven base units and none of them is for intelligence. That gap explains most of the confusion in AI progress claims.
2 min →Newsletter