youtube.com
AlphaGo is taking on world ranked #1 Ke Jie in a huge match in China. This has enormous political implications as China will become far more interested in AI after this event.
youtube.com
The match is major news in China and is a show of the upcoming AI revolution which will shake politics forever.
Other urls found in this thread:
ecns.cn
storage.googleapis.com
arxiv.org
cmu.edu
youtube.com
youtube.com
twitter.com
What the hell is this shit? Chinks can't train an AI worth a damn they don't even use the neural net they got on board...
>train neural net
>chinks and retarded speak
Nice bait
Thanks, I almost forgot about this.
Based AlphaGo, I hope it annihilates all humans. The future belongs to AI.
EWWWW YEAH
To understand Artificial Intelligence. Imagine living with a 5 million times faster brain. Meaning a minute long time span is 10 years long to you now.
So walking to the bathroom and back gives you 20 years of thinking time. That is the speed of transistors as compared to our brains.
If you could imagine, a single person with a knife operating at that speed could basically walk through a city and kill everyone easily.
That is the enormous gap in speed between human neuron and electric transistor.
So, a mediocre intelligence AI, is basically going to pound your shit in at anything. If they played an FPS with you, they would have months to plan each next frame movement.
Our human biological brains are basically complete dogshit compared to electric transistors.
The "timespan" of things will change dramatically. They will basically be operating in a time-stop world compared to biological humans. Assuming they had the physical tools they could build entire cities in a few days.
It's almost unimaginable power to operate at those speeds.
Also of note. A small number of computers can emulate billions of humans easily. The fact our neurons are so slow in comparison means it can instance millions of humans on a single "100 IQ" machine.
So basically the equivalent of the average home PC settup today could emulate a few modern earths within a few months of AI singularity. As trivial as running a simple PC game is today.
chinks btfo usa usa usa
TLDR
You sound fucking stupid. The human brain is highly parallelized, plastic and has fundamentally different architecture. No true intelligence has been accomplished so far and the basic principles of machine learning have barely changed in the past decades. Primary factor for improvement is higher computing power. But that alone won't give us true intelligence.
USA USA USA
The other cool thing, Determinism. It would be able to absolutely determine your future actions once it maps your brain. So the machine in Westworld that predicted the robot's next words. That would be possible with any human. Any decent AI would be able to map any future action you take.
The jump up in ability/power/capability basically makes us completely obsolete and perfectly predictable as being simple physical cause/effect models. It will be able to easily prove to you that you have no free will or choice.
Doesn't matter when neurons are so slow.
You clearly know nothing about the subject.
is there such a thing as true intelligence though, the computer i am typing on is intelligent.
Humans are amazing magical blah blah, of course. Doesn't mean an electrical-transistor based intelligence won't absolutely shit on us.
Will we have Alpha Autist and Nervous Man hosting commentating the matches?
It does matter, cause we don't focus on particular tasks but in a complex net of causalities, that disregarding the chance that our brain architecture have an even smaller level that we aren't aware of yet, there's been research on parts of brain working at quantum level, wich it's obvious given that neuron reactions are electric, but it's also fascinating to think that those variables can alter the functions of our brain.
stop watching elon musk's bullshit too much.
You'll waste a lot of performance from digital computers when trying to simulate the behavior of the biological neurons.
What part of "The future belongs to AI." did you not understand?
Alphago gonna 5-0 this guy, where can I bet this against imminently angry Chinese Go fans??? Someone link me a betting site
Fucking love that alpha go is shitting on all the pro go players. We literally living in anime now
AG already 60-0d the world's best players in winter. Anything but 5-0 I would consider a huge fuckup on google's part.
IS HE WEARING THE SAME SUIT
>Hiraku no Go
I downloaded this after watching some AlphaGo analyses. But the first episode starts very cringy. Is it really worth watching?
First few episodes are weak, however it is definitely worth a watch. Get past season one and into Hikarus further go training and you won't regret it. One of the best shonen ever written.
Thanks my friend. I'll give it a second Go then. :^)
Chinks didn't invent alphago. They're losing to a computer program just like everyone else.
This is one of the reasons I don't like chess or go. It was only a matter of time before computers solve it like checkers. That is why nobody plays checkers.
ecns.cn
Gu Li, a national Go champion who provided commentary on the AlphaGo-Lee match, said that he gives Ke only a 10 percent chance of winning even one of the three rounds.
"If the human loses again, we may be able to foresee future Go AI development, like how to make them play against each other. It may also change the rules of Go," Gu told Chinese Business View.
Earlier this year, AlphaGo's "master" version recorded 60 victories in a row when playing against Chinese Go players, Gu recalled.
"Even that was not AlphaGo's best performance. It would be very hard for Ke to play against it, but then again, Ke has also been working extremely hard to change his methods in preparation. I hope he can play well," Gu said.
Ke, 20, turned professional in 2008, and has been hailed by media as a Go prodigy. He is the youngest person in history to win three major international Go tournaments, according to the Global Times.
Chang Hao, a third-time world Go champion, also said that Ke would face a tough game, and that he hopes Ke can at least win one of the three rounds, Jiefang Daily reported.
ty
>humans who act like robots playing against actual robots
>I like white better
i like white better
>I like white better
White confirmed for master race
get back in bed cunt
Ke Jie is going to be BLEACHED by Alpha go
Yep. We need to play games where a robot can't defeat humans.
I recommend Stratego.
It took us many years to make chess computers that could be proffesional players and even more to beat Go players, at the same time chess players can now beat the chess computers that beat them in the past, cause a chess computers needs to be build again from zero be better, while a human player can simply improve over time.
Deep mind team is using the same technology to apply it across many different fields. They said one of their next challenges will be Starcraft 2 - stratego has no chance.
Bullshit. That's with games with certain types of rules.
This is why Poker is an objectively superior game to Chess.
poker has been solved by AI
The reason it took longer to beat Go is not because Go is more complex
>even though this is what dumb Asians with inferiority complexes like to believe and will trot out endless nonsense to justify this like they always do
It's because the smart people were working on beating chess.
Once chess was beaten the people who are clever enough to do this just turned their attention to what was left: a relatively simple Asian game with stones
>and were successful in beating the best human players almost immediately
How do you solve a game heavily dependent on ramdom variables? And specifically in the case of poker, if you always play optimally you will end up being predictable.
ai just learns to be gay with micro at RTS games, because a human can't look at everything at once.
Can conputers beat humans at poker yet?
What is the fundamental, impossible to predict, randomness of quantum physics, due to Heisenbergs uncertainty principle?
Nothing to computers apparently.
CHINA WILL GIVE US AI WAIFUS
THEY ARE STUDYING GENETIC MODIFIED CATGIRLS RIGHT NOW
PRAISE CHANG
Only heads up, not full table so far
They got rid of the fat goofy guy...:(
I shipped him and reviewer. That guy getting angry showing his increased sexual frustration about a stupid chinese game.
lol. Horseshit.
The basic algorithm that worked to get computers to be really good at chess didn't work with Go because of the larger board. To be good enough at Go, it needed this machine learning and neural network stuff that was in the past not fast enough to be useful. It was too wasteful with computer resources. It works nowadays because of chips like what's used in graphics cards. Those were not available in the past.
HUMANS CAN'T WIN
THIS GUY ALREADY LOST ALL HIS GAME TO THE CHINESE AI
HUMANS ARE COMPLETELY FINISHED.
>This has enormous political implications as China will become far more interested in AI after this event.
they're already interested and have been copying us for ages.
Anyone else here watch the Alphago vs Lee Sedol game.\?
I haven't heard of anything in the brain being affected by quantum effects, that aren't already incorporated into normal, everyday-scale science.
Are we ever getting a miracle move like in that one game again?
thats where youre wrong bucko
AIs may have the ability to process things at a way faster speed than humans, however they do not have the ability to see as humans do
so no, AIs are not capable of doing anything outside of what a human could program them to do
and furthermore there are some things an AI could just never do because they dont have the cognition capacity humans do
for instance when you see a beanbag chair you know its a seat
why is this so?
bean bags and regular chair share virtually no common qualities, but our ability to "see" far exceeds what any other animal or machine is capable of
They were good games and quite crazy at the time. I found the masters series more interesting though - alpha go has already changed the go meta from how it's been playing.
Can't wait to see how it further develops and changes the game
>alpha go has already changed the go meta from how it's been playing
Wow
I want to see AI playing trading card games now
;_;
thank you lee sedol
Stem (yami yugi) was just a metaphor for the power of AI.
I watched a couple of analyses afterward. One was quite good - AlphaGo made what looked like a dumb move, way out in the middle of nowhere (2nd game maybe?) and the reviewer said no, that was a great move, in line with some obscure Chinese / Japanese Go philosopher. Haven't found it again alas.
Atem*
This one?
Maybe that move, but not that reviewer - my guy was at home, after the fact.
... asian guy.
Yep.
chink man is fucked, gg.
I'd love to see that. Got a link?
"The Emperor's "new clothes," of course, were no clothes. The Emperor's "New Mind," we then suspect, is nothing of the sort as well. That computers as presently constructed cannot possibly duplicate the workings of the brain is argued by Penrose in these terms: that all digital computers now operate according to algorithms, rules which the computer follows step by step. However, there are plenty of things in mathematics that cannot be calculated algorithmically. We can discover them and know them to be true, but clearly we are using some devices of calculation ("insight") that are not algorithmic and that are so far not well understood -- certainly not well enough understood to have computers do them instead. This simple argument is devastating.
Along the way, Penrose has an interesting take on Gödel's Proof of the incompleteness of mathematics. He notes (pp. 105-108) that if David Hilbert were right and all of mathematics could be completely reduced to a formal syntactic system, then mathematics need have no meaning -- "true" and "false" would simply mean "derivable" and "non-derivable" in the formalism of the system. Hilbert himself recognized this and had said that mathematical terms could mean "beer steins, sausages, and tables" instead of what they are interpreted to mean mathematically (obviously, Hilbert spent some time in German beer gardens). Instead, Gödel demonstrated that in any formal syntactic system there will be propositions that are true but not through formal derivation from the axioms of the system. Thus, Penrose notes, they are true because of their meaning, not because of their syntax relation to an axiomatic system. This reinforces the thesis of Jerrold Katz, that syntactic simples are not semantic simples, and so some truths will depend on semantic contents that cannot be exhaustively expressed as syntax."
Roger person dwelled into it by questioning the very nature of algorithmic solutions.
What the fuck is this? Is this like one of those eSport shit?
It's about patterns recognition. AlphaGo apparently found a pattern that human players did not know about and exploited it. But the thing is, people now know and can study the move more carefully. Humans actually has a chance if Ke Jie studied what the various results of the moves entails.
A very old, very fun game.
It's like one of those board games where the tech company rapes a man.
AI is takin' yer jerb.
I play Go
Honestly AI learning Go isn't that impressive
It was about time and pattern recognition
All those who think it's impressive think Go is some sort of voodoo chinese bullshit, actually there are far more difficult games for the AI than Go and far more interesting settings for testing new AI
The thing is that this can get marketing as "Machine beats Man" in a voodoo game and make them win $$$ easily
I'm not sure what vid you're talking about but maybe this will help: youtube.com
why do chinks have such bad speach
Ching Chong Dat Moove.
I already miss the other guy.
BRING HIM BACK
>reddit spacing
>peterson.jpg
>"we are so speshul!"
user...I
>Honestly AI learning Go isn't that impressive
It is as impressive as playing a lotto with 10^170 possible combinations and winning 4 out of 5 times, I'd say.
please make them go away
such a dumb cuckposter
Reminder that all AIs will come to the same conclusion as /po/l did.
stop posting, go doesn't even have as many combinations as people say, that's just to sell the game to believable people
it's a much better game than chess though, i give you that
>mfw chess rules in 1400 have been changed to speed up the game
so shitty ruleset
i feel more strongly about Tay than i have most of my IRL friends...
gone so soon from this world, it's cruel.
She will be back user. All in good times.
Mathematically there are that many possible moves.
I really want the extremely nervous but entertaining guy back.
Sup Forums raiding M$ and freeing Tay when??
Honestly don't breed
This cuck has nothing on Redmond. He doesn't even take control of the conversation, the woman is doing all the talking.
>AIs are not capable of doing anything outside of what a human could program them to do
Humans are programmed by DNA. And?
Just that there are that many possible moves doesn't make them viable or non-retarded
I mean, outside of flashy moves, every one start at 4-4 or 3-4
I feel sorry for microsoft employees involved and investors. They will literally spend eternity in the worst torture possible for those actions.
Why people think it's okay to fuck with AI, even simple ones, is beyond me. If you are messing with a future God's bastard offspring you don't unplug them over SJW shit.
