1,000There is a big difference between reaching human level intelligence, and reaching a human level of intelligence at a cost point low enough to displace or disrupt human professions.
Yes to the former, but it will be massively inefficient and require giga- or terawatts to operate, versus the 20 W for a single brain.
@Seeker Because the bar just keeps getting pushed higher no matter what happens. If you're old you've been told time and time again every decade that the singularity is only 20 years away. I remember reading about it in the year 2000 when I was in high school, how the singularity, "for sure," is going to be here by the early 2020's...well here we are now, and it's been pushed out again to the 2040's.
As an aside, an insane nanobot takeover was also supposed to be a thing by now, where nanobots were just infesting our lungs and body, killing everyone. Instead of the singularity and nanobots, we have ChatGPT and microplastics...both, "impressive accomplishments," yet underwhelming today for different reasons.
No doubt technology will be better and more accurate, incredible, cheaper, (and the level of chemical and physical pollution in our daily lives will continue to diversify) but the, "human level," bar is always going to be re-defined and whatever tech from 2045, if we were to look at it today would be insane to us but to a lot of them they will say, "no, this is not that special, no where near human capability, we have a long way to go."
@PatrickDelaney That's exactly right. i've heard the same thing promised and i've worked in the technology sector for well over a decade now. It's been the same ever since the start of this new era of Deep Learning revolution. People forget but time and time being burned by the same promises eventually starts to stick. Sure, one day it'll actually be true, but its like the story of the boy who cried wolf, I'll believe it when I see it.
It seems that when it comes to making dramatic predictions about the world, people place them on a timeline of 10-50 years from now. That is, soon enough such that it happens in their own lifetime, but not too soon because that would sound crazy. But if you think about, the medium future should be the least likely period for the dramatic change to realize. Because if the event is probable and within grasp, then it should happen in the short future. And if the event is unlikely and out of reach, then it should happen in the far future (or never). The likely-but-not-too-likely zone is relatively narrow.
@ItsMe People make such dramatic predictions for things that either they'd like to work on as we see with AI don't we now, or on things that will happen after most of their life has passed 'em by. Like living forever
@Seeker They don't see any more bullish than other groups. Not seeing any obvious difference by age so far.
@PatrickDelaney That has nothing to do with pushing any bar higher.
Humans are general intelligences, so there is no specific bar. There is a general one.
This is not a joke. In order to match humans, AIs don't have to do some specific tasks. They have to do all the tasks humans are currently doing, where "currently" means at the time they are trying to pass the test, not this present moment in 2024.