>Another piece of evidence on the pile that the sudden panic and desire to "slow down" is because they're hitting the plateau on capability

if true then LLM related AI (post-post AI winter AI?) is probably one of the fastest inception-to-plateau tech sectors to have ever existed.

We're still improving transistors on a somewhat routine basis.

▲

mixdup 5 hours ago | parent [-]

The plateau doesn't have to be perfectly flat, but it's not a straight line upward anymore either (kind of like our work on transistors, where we've kind of hit the bounds of speed in clock cycles but are improving on miniaturization and power efficiency)

▲

password54321 5 hours ago | parent [-]

It took 6 years to solve ARC-AGI 1, 1 year to solve ARC-AGI 2 and 6 months to solve ARC-AGI 3.

▲

delillos 5 hours ago | parent [-]

Those version numbers don't necessarily correspond to equal increases in "difficulty", though.

	▲	password54321 5 hours ago \| parent \| next [-]
		Correct, the benchmark became exponentially more difficult as it progressed from pattern matching puzzles to games.
	▲	JacobAsmuth 4 hours ago \| parent \| prev [-]
		Very true, the sharp increase in difficulty (as measured by human passrate plummeting from 1->2 and again from 2->3) gives an even more stark view of AI capabilities over time.