Humans can’t beat AI at Go, aside from these exploits
kek, reminds me of when I was a wee one and I'd 0 to death chain grab someone in smash bros. The lads would cry and gnash their teeth about how I was only winning b.c. of exploits. My response? Just don't get grabbed. I'd advise "superhuman" Go systems to do the same. Don't want to get cheesed out of a W? Then don't use a strat that's easily countered by monkey brains. And as far as designing an adversarial system to find these 'exploits', who the hell cares? There's no magic barrier between internalized and externalized cognition.
This is literally the dumbest shit I've read all week and it's been a pretty dumb week. I'm afraid I have to diagnose Roko with having the brain scamblies. There is no cure.
In my more innocent college days, there was a group of people doing a reading of it in the dorm lounge laughing their asses off. Ong I thought it was a hyper self-aware satire that was making fun of internet "umm ackshully" / iamverysmart posters. There's no way someone earnestly spent their time writing over half a million words on a self-insert Harry Potter fanfic as some form of mental masturbation... right?
In chess the table base for optimal moves with only 7 pieces takes like ~20 terrabytes to store. And in that DB there are bizzare checkmates that take 100 + moves even with perfect precision- ignoring the 50 move rule. I wonder if the reason these adversarial strats exists is because whatever the policy network/value network learns is way, way smaller than the minimum size of the "true" position eval function for Go. Thus you'll just invariably get these counter play attacks as compression artifacts.
Ah, I see TWG made the rookie mistake of thinking they could endear themselves to internet bigots by carrying water for them.
^Also, fuck this nazi infested shithole. Absolute eye bleach.
Wishful thinking on my part to think their sexism/eugenics posting was based on ignorance instead of deliberately being massive piles of shit. Don't let them know Iceland has the highest number of GMs/pop or else we'll get a 10,000 page essay about how proximity to volcanoes gives + 20 IQ points.
Edit: Judit Polgár for ref if anyone wants to learn about one of the greatest of all times. Her dad claimed he was doing a nature/nurture experiment in order to prove that anyone could be great if they were trained to master a skill from a young age, so taught his 3 daughters chess. Judit achieved the rank of number 8 in the world OVERALL and beat multiple WC including Kasparov over her career.
idk its almost like if more girls were encouraged to play chess and felt welcome in the community these apparent skill differences might disappear
Without fail in the comments section, we have Daniel Kokotajlo (the philosophy student turned ai safety advocate who recently got canned at OAI) making the claim that "we [ = Young Daniel and our olde friend Big Yud] are AI experts and believe that risking full scale nuclear war over data centers is actually highly rational^{tm}" :)
...anyways, what were we saying about David Gerard being a bad faith actor again?
Might be slightly off topic, but interesting result using adversarial strategies against RL trained Go machines.
Quote: Humans able use the adversarial bots’ tactics to beat expert Go AI systems, does it still make sense to call those systems superhuman? “It’s a great question I definitely wrestled with,” Gleave says. “We’ve started saying ‘typically superhuman’.” David Wu, a computer scientist in New York City who first developed KataGo, says strong Go AIs are “superhuman on average” but not “superhuman in the worst cases”.
Me thinks the AI bros jumped the gun a little too early declaring victory on this one.
The doom prediction in question? Dec 31st 2024. It's been an honour serving with you lads. 🫡
Edit: as a super forecastor, my P(Connor will shut the fuck up due to being catastrophically wrong | I wake up on Jan 1st with a pounding hang over) = (1/10)^100
Every day I become more convinced that this acct is an elaborate psyop being run by Yann LeCun to discredit doomers. Nobody could be this gullible IRL, right?
Super weird that despite receiving 20 million dollars in funding from SBF & co. and not being able to shut the fuck up about 10^^^10 future human lives the moment he goes on a podcast, Danny boy insists that any allegations that he is lobbying on behalf of the EAs are simply preposterous.
Now please hand over your gpus uwu, it’s for your safety 🤗 we don’t allow people to have fissile material, so why would we allow them to multiply matrices?
BigMuffin69 @ BigMuffin69 @awful.systems Posts 8Comments 247Joined 2 yr. ago
kek, reminds me of when I was a wee one and I'd 0 to death chain grab someone in smash bros. The lads would cry and gnash their teeth about how I was only winning b.c. of exploits. My response? Just don't get grabbed. I'd advise "superhuman" Go systems to do the same. Don't want to get cheesed out of a W? Then don't use a strat that's easily countered by monkey brains. And as far as designing an adversarial system to find these 'exploits', who the hell cares? There's no magic barrier between internalized and externalized cognition.
Just get good bruv.