3 ms·
Ambitious projects and optimistic predictions have seemed not-unusual among big-name AI people, since the start. Marvin Minsky predicted emergent behavior and
by neilv 7y ago
Ambitious projects and optimistic predictions have seemed not-unusual among big-name AI people, since the start.
Marvin Minsky predicted emergent behavior and recursive effects much sooner. (I think this was in all honesty: not realizing how complex the human brain, and not knowing how hardware would develop.)
Douglas Lenat apparently thought commonsense knowledge and reasoning would be easier.
Ray Kurzweil was often making predictions about hard-AI accomplishments just around the corner, which we're starting to see much later.
- ghostbrainalpha 7y agoInteresting note about Ray Kurzweil, his predictions on when humanity accomplishes human brain simulation don't predict we will achieve this until the next decade. (2020-2030)
- echelon 7y agoWhich is still way too early. Everything we thought was just around the corner is not: flying cars, self-driving cars, AGI, cold fusion, the singularity... Hype cycles are a thing. I for one don't think we'll see any of these in our lifetime.
- tim333 7y agoGood self driving cars maybe but if you don't mind an iffy one here you go: https://www.youtube.com/watch?v=aaOB-ErYq6Y https://www.youtube.com/watch?v=aaOB-ErYq6Y
- tim333 7y agoWhile Kurzweil is cranky in many areas like his 100 pills a day, his extrapolations of Moore's law type growth patterns seem quite sensible - here's the log graph, let's get a ruler and extrapolate the pattern continuing. Not really rocket science and you don't need Kurzweil to do it but he's perhaps the leading populizer of the approach.
- Chathamization 7y agoKurzweil is an interesting case. His predictions have been impressively wrong for the most part, yet he seems to refuse to admit it.