>>12209709It'd eventually be much faster than humans, and it's definitely possible if you're setting no time limit and assume no apocalyptic event, be it 50/100/500 years.
It's just not that practical to talk about so soon before it actually exists.
>>12209935This shows we're far away from something totally groundbreaking. I think there's a decent chance we may end up kicking the can down the road due to failure to fully understand and replicate neurological systems, and instead get there with a different approach. And then once that gets good enough, the AGI could possibly help us go back and understand neuroscience better. Then maybe insights and ideas from that could complement the existing techniques.