@Cyberstalker's comment on 'A frickton of AI stuff out of the blue ( Shit's unironically crazy right now )'

Humanity figured out super advanced better than all humans narrow AI variant years ago. We don't really need to figure out AGI to advance humanity at unprecedented rates, we could do the same thing by creating a thousand different narrow AIs specializing in different fields. If humanity is really at that point where we can make self optimizing and self improving factories, then there really isn't much else to be done. Technology is already halfway out of human hands. We are the midwives of the future. An expert intelligence AI would be truly advantageous to the species as you would suddenly have all of humanity having access to all the knowledge of the world up to a decade or two with a qualified teacher to teach it well.

Also, if China fails to surpass the US ( Very unlikely ), India will definitely surpass the US, because unlike China, India will have China and South East Asia to fund its markets to surpass the western markets.

As expected, Elon Musk's net worth will cross a trillion USD with neuralink and SpaceX.

Conclusion:

The future is here, and it is moving at faster rates than even us nerds can track anymore.

Jump in the discussion.

No email address required.

View entire discussion

Cyberstalker he/him Aspiring hikikomori 4mo ago #6674165

This diagram is extremely mid-witted and originates from the assumption that AI progresses towards being more humanlike no matter what it's purpose is, it's very possible for 2-5 to be done without any understanding of communication, and 3 does not require 2.

42 Context

gigachad_brony they/them that's how you get the hobo stole my baby story Cyberstalker 4mo ago #6674284

Explain further

7 Context

Cyberstalker he/him Aspiring hikikomori gigachad_brony 4mo ago #6676488

Movie-brained redditors think that it's inevitable for a powerful enough AI system that manages data to not only become sentient, but have humanlike motivations (DAE le skynet??? The AI we trained to act like samples of text said yes when we asked if it was a person, we're doomed!). It's also possible to make a very specific AI system that takes actions in one scenario but does not have human-like generalization, and does not approach the problem in the same way as a human does even if it ends up with a similar solution after a large number of iterations (many AI models are very good at a single scenario, but cannot extrapolate outside of it).

2 Context

gigachad_brony they/them that's how you get the hobo stole my baby story Cyberstalker 4mo ago #6676504

This argument is as stupid as saying whether String theory is correct or not. There simply isn't enough data and methods to confirm the idea one way or the other. You are complaining that the unitary intelligence people are wrong meanwhile the multiple types of intelligence people are right.

1 Context

Cyberstalker he/him Aspiring hikikomori gigachad_brony 4mo ago #6676607

Ok, I guess "the unitary intelligence people are wrong meanwhile the multiple types of intelligence people are right." is just my opinion with regards to AI

3 Context

gigachad_brony they/them that's how you get the hobo stole my baby story Cyberstalker 4mo ago #6676630

That's fair. We had a good conversation.

Cyberstalker he/him Aspiring hikikomori gigachad_brony 4mo ago #6676791

I wasn't trying to argue with you, I was arguing with the image

gigachad_brony they/them that's how you get the hobo stole my baby story Cyberstalker 4mo ago #6677062

Cool. We still had a good conversation.

fartnigga they/them Getting downvoted for this comment is making me feel reeeeeeal uncomfortable gigachad_brony 4mo ago #6674400

why would it need to be able to take actions before it can innovate? why would it need to innovate before it can organize? most HR workers haven't achieved levels 2 or 4, which is why they're assigned to level 5