4 ms·
That poll is interesting, but IMHO shows the implicit meaninglessness of predictions to 10/20 years. Do you feel we've done 30% of the work towards human level
by riffraff 10y ago
That poll is interesting, but IMHO shows the implicit meaninglessness of predictions to 10/20 years.
Do you feel we've done 30% of the work towards human level AI between 2014 and 2017, given a target of 2024?
- nbouscal 10y agoWe went from not even beating good amateurs at go to beating the best pros in the world. I don't think it's like a progress bar we should be expecting to fill at a linear pace, but, that certainly seems like a strong indicator we've come a long way (and it's one of many examples I could have chosen).
- riffraff 10y agoit certainly isn't a linear bar, but any improvement in AI in the last three years does not look at all as getting closer to human level intelligence. We have just become better at some tasks that we knew how to automate for a long time, albeit badly. We don't have a clue on how human intelligence is supposed to work.