Ajeya Cotra

Since Kotra and Davidson were expecting AI to get 3.6 times better every year, but it actually got 10.7 times better every year, it's no mystery why their timelines were off.

When John recalculates Davidson's model with Epic's numbers, he finds that it estimates AGI in 2030, which matches the current vibes.

4.

With this information in place, it's worth looking at some prominent contemporaneous critiques of bio-anchors.

Various people criticized bio-anchors, many strange anchors, for how much compute it would take to produce AGI.

For example, one anchor estimated that it would take 10 to the power of 45 flops, because that's how many calculations happened in all the brains of all animals throughout the evolutionary history, which eventually produced the human brain that AIs are trying to imitate.

To make things even weirder,

This anchor assumed away all other animals other than nematodes as a rounding error.

Fact check, true.

All of these seem to detract from the main show, an attempt to estimate the compute involved in the human brain.

But even this more sober anchor was complicated by time horizons.

It's not enough to imitate the human brain for one second, AIs need to be able to imitate the human brain's capacity for long-term planning.

Kotra calculated how much compute AGI would require if it needed a planning horizon of seconds, weeks, or years.

Thanks to META, we now know that existing AIs have already passed a point where they can do most tasks that take humans seconds, are moving through the hour range, and are just about to touch one day.

So the seconds anchor is ruled out.

But it also seems unlikely that AGI will require years, because most human projects don't take years, or at least can be split into tasks that take less than one year each.

Intuition pump.

Are we sure the average employee stays at an AI lab for more than a year?

If not, that proves that a chain of people with sub-one-year time horizons can do valuable work.

The AI Futures team guessed that the time horizon necessary for AIs to really start serious recursive self-improvement was between a few weeks and a few months, though this might look like a totally different number on the meter graph, which doesn't translate perfectly into real life.