4 ms·

If it’s not superintelligent now, it will be within 5 years. Probably less.

• _c3ag 3y ago
  i love how humans (we, btw) guesses stuff with numbers! first time i thought about it was when a friend started to talk about % of stuff of himself... i was: tf did you came with these numbers
• gizajob 3y ago
  This statement is now about 70 years old.
  • maxdoop 3y ago
    Is there any progress in AI that you find significant or what do you think would qualify as a “breakthrough”?
    • gizajob 3y ago
      There’s been heaps of breakthroughs and loads of significant progress. But I don’t think there’s likely to be a “singularity” type event any time soon. Plus all the major breakthroughs in AI haven’t really had the disastrous consequences predicted beforehand, and I think increasing computer intelligence is likely to be similar. Deep Blue beating Kasparov didn’t destroy chess, it served to make better and better human chess players. We also rapidly evolve socially to integrate better and better machines, and AI research has the implicit assumption that we stay still while the machine leaps and it’s not like that at all, like a question one would have to ask nowadays of a Turing-type-test, even a layman, would be “what if it’s just ChatGPT answering…”