> a student who gets 100% on a problem set in 10 minutes is more intelligent than one with the same score after 120 minutes

According to my mathematical model, the faster student would have higher effectiveness, not necessarily higher intelligence. Resource consumption and speed are practical technological concerns, but they're irrelevant in a theorical conceptualization of intelligence.

If you disregard time, all computers have maximal intelligence, they can enumerate all programs and compute answers to any decidable question.

Yeah speed is a key factor in intelligence. And actually one of the biggest differentiators in human iq measurements

Humans are a bit annoying that way, because it's all correlated.

So a human with a better response time, also tends to give you more intelligent answers, even when time is not a factor.

For a computer, you can arbitrarily slow them down (or speed them up), and still get the same answer.

Maybe. If I could ask a AI to come up with a 50% efficient mass market solar panel, I don’t really care if it takes a few weeks or a year if it can solve that though. I’m not sure if inventiveness or novelness of solution could be a metric. I suppose that is superintelligence rather than AGI? And by then there would be no question of what it is

> response time

Imagine you take an extraordinarily smart person, and put them on a fast spaceship that causes time dilation.

Does that mean that they are stupider while in transit, and they regain their intelligence when it slows down?

Who is a better free-thrower, someone who can hit 20 free throws per minute on Earth, or the same thrower who logged 20 million free throws in the apparent two years he was gone but comes back ready for retirement?

No, because intelligence is relative to your local context.

Why should one kind of phenomenon which slows down performance on the test be given a special "you're more intelligent than you seem" exception, but not others?

If we are required to break the seal on the black-box and investigate the exactly how the agent is operating in order to judge its "intelligence"... Doesn't that kinda ruin the up-thread stuff about judging with equations?

The point is that the wall-clock measuring the test-taker's performance should be in the same context as the test-taker.

They will still complete the task in 70 local minutes, even if that's eighty of an outside observer's.