5 ms·
Big tech’s quest for human-level A.I
- neonate 7y agohttp://archive.md/VSgOL http://archive.md/VSgOL
- siberianbear 7y agoThank you!
- m0zg 7y agoI wish they'd knock it off with the grandiose headlines. I make my living with what passes for AI nowadays, and I can say with complete certainty that we will not be achieving anything anywhere close to general AI in my lifetime. I'd be willing to bet serious money on that, too, if there were a way for me to collect. Literally nobody is working on human-level AI. There are no papers, no conferences, nothing. And yet, we see these headlines with persistence deserving of a better cause.
- the8472 7y ago> Literally nobody is working on human-level AI. As a concrete project? True. As a long term vision? False. OpenAI and MIRI were explicitly founded with AGI doomsday scenarios in mind and how to do fundamental AI research to avert those by making sure that when we reach human level AI eventually that it'll be safe. The blue brain project exists too. Of course nobody will say that they'll have a timeline or even a concrete plan build an AGI in the same way that nobody will say they have a cure for cancer or HIV today or tomorrow even though the hope behind research is to have effective cures one day. > I make my living with what passes for AI nowaday Many practical applications are not exactly bleeding edge research or even applications of the latest research results. Perceptrons are from the 50s and still have applications today, AMD's branch predictor uses them for example. That doesn't negate the fact that the field advances.