I forgot that most people are not so surprisingly unafraid of AI because they somehow think it is not dangerous to share the planet with a species that is more intelligent than your own, but because they still cannot imagine that it will happen.
-
-
Replying to @Plinz
I'm just not at all convinced intelligence works like this implies it does - that is, the smartest person wins, and everyone else stands around being useless - rather than it aggregating almost linearly. I just don't know. How do you know? (Do you know?)
2 replies 0 retweets 0 likes -
That only holds if you confine your comparisons to human individuals. Within humanity, the smartest person indeed doesn't always win. Within the global biosphere, the smartest SPECIES has utterly cleaned house.
1 reply 0 retweets 0 likes -
Replying to @NapalmSushi @jprwg
From the perspective of every other primate species, human intelligence is an uncontrollable weapon of mass destruction. Under which conditions should we treat the research efforts into superhuman artificial intelligence as careless experimentation with an extinction risk?
1 reply 0 retweets 0 likes -
Replying to @Plinz @NapalmSushi
Ok, but I'm not a primate species, I'm a primate, as are you. From my perspective, almost the entire world is out of my control. The question is: what does 1000x human intelligence imply? More "can dominate humanity 1000x over" or "can substitute for 1000x humans"?
2 replies 0 retweets 0 likes -
Replying to @jprwg @NapalmSushi
Our intelligence does not scale well, because brains cannot grow much larger, childhoods (= initial training periods) cannot be much extended, and communication between minds is limited. There is no obvious comparable limit for an electronic brain. One may outsmart all of us.
1 reply 0 retweets 0 likes -
Replying to @Plinz @NapalmSushi
But what does "outsmart" imply?? Are you sure it implies what you're implying it implies?
1 reply 0 retweets 0 likes -
The image "outsmart" suggests to me is a line of humans all queuing up to try to beat the superintelligence at a game of chess or similar. "Damn, that was out smartest guy, & he was still hopelessly outmatched!" I'm not not at all convinced that's how aggregate capability works.
2 replies 0 retweets 0 likes -
Replying to @jprwg @NapalmSushi
Imagine that one of the first scalable general problem solvers is a corporate intelligence that is tasked with gaming financial markets (what we euphemistically call "trading"), it might turn out that there is no way we can stop it before it totally obliterates the world economy.
1 reply 0 retweets 0 likes -
Replying to @Plinz @NapalmSushi
Uh, maybe? How do returns to financial trading work today? Is it the case that the team w/ the smartest person gets most of the gains & the rest get little? If not, why wld the AI do better than a team(s) of people equal in no. to the no. of times smarter than human the AI is?
1 reply 0 retweets 0 likes
Trading gains depend on exploiting arbitrage opportunities. The best arbitrary opportunities depend on zero day attacks on existing financial ecosystems, like Soros hacking the pound. If you are a hedge fund, you can reinvest your gains into bidding for the best quants.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.