5 ms·

Ten years from now will either be: a) Remember all that fuss about AI destroying the world? Lol. ~ or ~ b) I'm so glad those people stepped in to save us fro

by tcgv 2y ago

Ten years from now will either be: a) Remember all that fuss about AI destroying the world? Lol. ~ or ~ b) I'm so glad those people stepped in to save us from doom! Which one do you think is more likely?

cosmic_quanta 2y ago
Unless AI starts being 1 000 000x energy efficient, my money is on a). The amount of energy required for AI to be dangerous to its creators is so vast that I can't see how it can realistically happen.
- whimsicalism 2y ago
  We know that we can run human level intelligence with relative efficiency. Without discussing timelines, it seems obvious that human energy usage should be an upper bound on the best possible energy efficiency of intelligence.
- kjkjadksj 2y ago
  That depends on how its used. See the terminator movies. One false positive is enough to end the world with even current AI tech if its merely mated to a nuclear arsenal (even a small one might see a global escalation). There have been false positives before, and the only reason why they didn't end in nuclear Armageddon was because the actual operators hesitated and defied standard protocol, which probably would have lead to the end the world as we know it.
- dindobre 2y ago
  If we manage to harness the ego energy transpiring from some people working on "AI" we should be halfway there!
its_ethan 2y ago
I'll bite... "a"
whimsicalism 2y ago
It will never be B even if the “safetyists” are correct. We rarely notice the near catastrophic misses except in obvious cases where we accidentally drop a nuke or something.
ALittleLight 2y ago
Or: c)
- tcgv 2y ago
  Fair enough! c) Humanity unleashed AI superintelligence, but safeguards proved inadequate, leading to our extinction
  - CRConrad 2y ago
    There, finally all the alternatives. So, doesn't that show that your original question was meaningless? Because, if there is any non-zero probability that the correct answer is c), should anyone really fuck around with this shit, regardless of how small the probability is?
    - tcgv 2y ago
      It's a good argument. However, Pandora's box has already been opened, and people will keep experimenting with this technology, there's no going back. A key question is how long it will take to achieve such dangerous AI: one year, ten years, or a hundred years? Ilya and his colleagues believe it will happen sooner rather than later and want to take action now. So the original question remains, since it's more about the timeframe rather than whether it will happen or not

by tcgv 2y ago

cosmic_quanta 2y ago

Unless AI starts being 1 000 000x energy efficient, my money is on a). The amount of energy required for AI to be dangerous to its creators is so vast that I can't see how it can realistically happen.

whimsicalism 2y ago
We know that we can run human level intelligence with relative efficiency. Without discussing timelines, it seems obvious that human energy usage should be an upper bound on the best possible energy efficiency of intelligence.
kjkjadksj 2y ago
That depends on how its used. See the terminator movies. One false positive is enough to end the world with even current AI tech if its merely mated to a nuclear arsenal (even a small one might see a global escalation). There have been false positives before, and the only reason why they didn't end in nuclear Armageddon was because the actual operators hesitated and defied standard protocol, which probably would have lead to the end the world as we know it.
dindobre 2y ago
If we manage to harness the ego energy transpiring from some people working on "AI" we should be halfway there!

its_ethan 2y ago

I'll bite... "a"

whimsicalism 2y ago

It will never be B even if the “safetyists” are correct. We rarely notice the near catastrophic misses except in obvious cases where we accidentally drop a nuke or something.

ALittleLight 2y ago

Or: c)

tcgv 2y ago
Fair enough! c) Humanity unleashed AI superintelligence, but safeguards proved inadequate, leading to our extinction
- CRConrad 2y ago
  There, finally all the alternatives. So, doesn't that show that your original question was meaningless? Because, if there is any non-zero probability that the correct answer is c), should anyone really fuck around with this shit, regardless of how small the probability is?
  - tcgv 2y ago
    It's a good argument. However, Pandora's box has already been opened, and people will keep experimenting with this technology, there's no going back. A key question is how long it will take to achieve such dangerous AI: one year, ten years, or a hundred years? Ilya and his colleagues believe it will happen sooner rather than later and want to take action now. So the original question remains, since it's more about the timeframe rather than whether it will happen or not