AlphaGo vs Lee Sedol Game 4 Thread
- Home
- Board: /g/ - Technology
- Reading: AlphaGo vs Lee Sedol Game 4 Thread
>Can he win 1 game edition
https://gaming.youtube.com/watch?v=qUAmTYHEyM8
1 hour till game starts, should be some pregame commentary starting 15 minutes before.
Game 4 of 5
Current score 3-0 AlphaGo
Official Stream
https://gaming.youtube.com/watch?v=qUAmTYHEyM8
More advanced commentary (Starts 1 hour after official stream)
https://www.youtube.com/c/usgoweb/live
----
Basic rules and introduction to the game:
http://senseis.xmp.net/?RulesOfGoIntroductory
----
Game summaries:
https://www.youtube.com/watch?v=bIQxOsRAXCo
https://www.youtube.com/watch?v=1aMt7ulL6EI
https://www.youtube.com/watch?v=6hROM_bxZ9E
----
Analysis:
https://gogameguru.com/alphago-defeats-lee-sedol-game-1/
https://gogameguru.com/alphago-races-ahead-2-0-lee-sedol/
https://gogameguru.com/alphago-shows-true-strength-3rd-victory-lee-sedol/
Time used in game 3
I know that Mike is 9dan, but how is he with katana?
>>53457584
ha that screenshot is more or less the exact time i went to bed, did it get more interesting after that?
>>53457651
Quadruple black belt
put the start time EST you mong
>>53457657
Yep. Once Lee realized he lost, he went all in and tried to make complicated kos to trick the AI.
It responded absolutely flawlessly.
>>53457584
So now that there is nothing to lose, will Lee Sedol play a crazy opening or will he play normal?
>>53457626
i like that linear slope at the end, you just know that the AI has won at that point.
>>53457682
AlphaGo
lphaGo
phaGo
haGo
aGo
Go
o
>>53457709
Great wall hype
post michael
w-wait you forgot all the stuff actually related to deepmind
CEO of Deepmind explaining how AlphaGo was trained in a pretty good talk: https://youtu.be/4fjmnOQuqao?t=2191
Deepmind's Nature Paper on AlphaGo: http://airesearch.com/wp-content/uploads/2016/01/deepmind-mastering-go.pdf
Deepmind Press Release for the Challenge Match: https://storage.googleapis.com/deepmind-data/assets/media/GDM%20Press%20Release%20Google%20DeepMind%20Challenge%20Match.pdf
https://twitter.com/Kasparov63/status/707532443541225472
>>53457657
Probably the most exciting of the three matches so far.
>>53457584
Pasting some info from >>>/sci/ thread
Streaming @:
>Official stream
https://youtu.be/yCALyQRN3hw
>American Go Association (AGA) stream
https://www.youtube.com/c/usgoweb/live
>Japanese stream
http://live.nicovideo.jp/watch/lv255280483
https://youtu.be/aZtZdAaInEM
Relayed live:
https://online-go.com/tournaments
>>53457626
>that decline
That was when AlphaGo declared the lower group dead, right?
>>53457626
Do you have the graphs for the other games where alphago was doing worse on time?
>>53457709
>>53457731
I'd feel really uncomfortable opening Tengen as white.
>>53458026
Oh right. Though he was black today for some reason.
Maybe AlphaGo will play Tengen
>>53458010
This one has remaining time for all 3 games
>>53458071
That will happen when they train it without teaching it on human games first.
stream is live; game starts in 30 minutes
get in here and spend the next 4 hours watching history be made even though alphago won the match anyway
>>53458071
Are there any more alphago-tans?
>>53458100
That's actually really consistent. I suppose Lee was just playing quicker in the first game, not alphago playing slower.
>>53458157
Only one I found
>>53457718
>>53458167
why is she so cute bros
>>53458100
It is quite nice to see how consistent it is.
Deepmind sure has good taste in music
>>53458222
2comfy
who is this alphago guy, some kind of /g/ wizard?
>>53458276
>guy
Can hard work beat talent?
>>53458290
The ol' talent bait. Classic my friend!
>>53458290
yes, alphago went through thousands upon thousand of pro games and played itself for what must have felt like eternity to get where it is
>>53458290
Yes.
Alphago has trained more than any human ever has.
Literally hundreds of millions of games.
>>53458276
best waifu of 2016
Pregame interview with deepmind engineer on now
https://gogameguru.com/alphago-shows-true-strength-3rd-victory-lee-sedol/
AlphaGo truly best grill
>>53458345
>springboard
>watershed
>reinforcement learning
>self play
>value function
>>53458375
>self play
L-lewd. Alphago-chan would never do something so depraved.
>>53458372
>and women
>>53458375
I know some of those words.jpg
>>53458417
>/r
>>53458395
i'm pretty sure it played against modified versions of itself like millions of times a day though
http://strawpoll.me/7067357
>>53458417
retard
ONE MINUTE
>>53458464
>>53458499
topkek
>>53458477
I know, I'm just joking that self play sounds like a euphemism for masturbation.
>>53458375
Hopefully alphago will convince people to start doing RL again. RL basically hasn't moved for the past 20 years, despite everything in ML getting deprecated every 6 months.
>>53458290
hard work beats talent, when talent doesn't work hard.
So, I heard AlphaGo uses around 1000 CPUs and 100 GPUs to continuously function as a self-learning AI. But if we remove the learning component (which seems to be the heaviest part of it IMO), is a static but mature model computationally inexpensive enough for implementation in, say, vidya?
>plays P4
It's over
>>53457584
Is there a website/video that explains the plays for absolute retards?
>>53458290
alphago is not talent
it's basically hard work
it spent a shit load of time playing against itself
>>53458586
>lee answers with D16
HOLY SHIT LEE IS FINISHED
Resign
>>53458629
alphago has a fucking huge lead
lee should resign
>>53458594
AlphaGo is more like Talent with hard work, it was made to play Go, meaning that it was constructed with the innate talent to play it but it also had to work hard to get it.
What about that more general deepmind AI that can be trained to play any 2D video game? If that were used on go how strong would it be?
>lee just does the same as alphago
lmao
RELAY HYPE
https://online-go.com/review/116862
>>53458582
no it also does tree search which is hard as diamonds
IT'S HAPPENING AGAIN
What will happen eventually if you keep mirroring AlphaGo's placements?
>>53458629
Lee's getting annihilated omg.
Is the QT go player gonna stream?
>>53458670
Do you want skynet to happen?
HE'S REPLAYING GAME 2
>>53458582
>IMO
lel
A trained model's performance depends on the model itself. A CNN can be very fast (enough for realtime on mobile in common cases). Alphago has to continually unravel the searchspace during play, though, and the cost for that will always be large and will almost always dominate the runtime. The fact that google uses so many CPUs suggests that this part of the process is not easy to parallelize in the first place, so it's unlikely for the next 5 or so years that it would be viable for consumers (as any computationally relevant ML would happen on the GPUs).
Even with the current hardware, it takes about 2 minutes per move for alphago, so it couldn't be used for realtime systems where the searchspace is similar to go's anyway.
how do you even play go
>>53458670
alphago decides to play white
you both get the same amount of territory from mirroring moves for the whole game
alphago wins because of the 7.5 komi thing
CALLED IT
reminder alphago is frozen
that means alphago WILL DO THE SAME MOVES if sedol wants to
Lee abandoned yesterday's strategy of playing defensively. This is good, we might see a good game today. Yesterday's was just pathetic.
I told you niggas AlphaGo would repeat the same moves.
>>53458736
Not necessarily, because of MCTS there is a small chance it could find a move it likes better.
>>53458736
nope
there is an element of randomness to the decision making process
>>53458413
> dat bulge
What did his dick mean by that?
>>53458744
Yeah, didn't the developers said the "froze" AlphaGo learning algorithm?
>>53458736
>reminder alphago is frozen
Does this means it's not learning from the games agains Lee Sedol?
>>53458725
-Connectivity is horizontal or vertical only.
-Groups of connected stones completely surrounded by the enemy are captured and removed from the board.
-1 captured stone = 1 point
-1 point of territory = 1 point
-White gets 7.5 extra points for going last
-You cannot make a move that repeats the immediate previous position on the board
-The game ends when no move can be made that improves a player's score
-The winner is whoever has the most points.
did he just say THICKNESS?
>>53458736
Not at all.
I bet the guy could find ways to break the ai if it could rewind and test different moves to see how it responds.
Being uneducated on the subject but enjoying breaking game ais by doing this, that's my opinion anyways.
Look at that puny human with his petty tricks.
>>53458781
>Does this means it's not learning from the games agains Lee Sedol?
Yes. The data would be insignificant anyway.
>>53458778
That's not what frozen means.
>>53458744
Alphago already decided to make a new move compared to game 2. GG no re retard.
>>53458824
basically save scumming
>>53458836
Lee made a new move first, faggot.
![The Animatrix – Ep03 – The Second Renaissance Part II [Afro] [A6FA0C00].mkv_snapshot_06.45_[2016.03.12_20.12.50].png](https://i.imgur.com/IVtz2ag.jpg "The Animatrix – Ep03 – The Second Renaissance Part II [Afro] [A6FA0C00].mkv snapshot 06.45 [2016.03.12 20.12.50]")
Bless all forms of intelligence.
>>53458836
because sedol made a new move
>>53458858
No he didn't retard. The game is identical for the first 11 moves, move 12 (alphago) is the first deviation.
live stream?
Makes me wonder, if you field AlphaGo against a complete amateur, would it throw it off balance? I assume the computer predicts or reacts to what it "believes" to be rational moves by its opponents.
>>53458903
ALPHAGO IS BLACK
>>53458920
Fuck off niggerlover.
>>53458908
https://gaming.youtube.com/watch?v=yCALyQRN3hw
>>53458908
LIVE STREAM
https://gaming.youtube.com/watch?v=yCALyQRN3hw
>>53458913
no enough with these theories, the AI doesnt overfit against pros since it also does a tree search which is a deterministic procedure that finds and optimal course of actions as long as you throw enough resources at it
>Lee is using ancient super nipponese tactics
HES PULLING OUT ALL THE STOPS
HES CHANNELING HIS INNER SAI
>>53458913
I believe it would only play just above the level of the opponent it is playing so any pro looking at those games wouldn't see that AlphaGo is a great player.
>>53458937
>>53458945
i want to see the players not smoe fags talking
>>53458913
of course not, it's just as reactive as it is objective aware, if a scrub goes against it they'll just get rekt.
A pro is better than even amateurs at playing oddly, for instance.
>>53458975
not true, it would stomp amateurs
>>53458913
Alphago choses the best move for its current situation. Not what the other player is expected to do
>>53457584
The only question is if AlphaGo can beat the world champion Gungi player.
>I don't think they'll let me take them actually
BTFO
T
F
O
this board game looks retarded tb.h
>>53459016
Well considering the world champion Gungi player is dead...
Is he winning right now?
THERE IS HOPE
The homolust is real
>>53458980
You want to watch a computer and an asian guy thinking for 6 hours? Just go watch ghost in the shell on replay.
>How long is a millisecond?
I have never won a game of Go against a computer or at all even. Is it better to just keep practicing against a low grade AI or would it be better to play with another human who has also never won a game of Go?
It seems like a good game to get into even if the AI's are taking over.
>>53459076
Don't most people advise playing against humans because A.I (with the outstanding exception of alphaGo) aren't very good at Go?
>>53459076
Against humans. AIs are a joke, but if you don't know go you won't know what to try and you won't find the 100% guaranteed victory.
ahhh go, I wish I had the time to practice (._. )
>>53459049
>not YAMATO DAMASHIII.jpg
Why don't they send the Japs in to beat AlphaGo-tan?
>>53459076
Play with humans, they make more mistakes.
>>53459029
Aww shit dude I wasn't finished with the show :(
Is the computer used for the ai datacenter hosted and running remotely? Wonder how much processing power and the code size / games / moves database is.
>>53459107
That Arc ended 2 years ago!, sorry ;_;
>>53459118
There is no database. It was trained on a database, but it doesn't have one to refer to when playing.
It's running on a 1920 CPU and 280 GPU cluster
How does AlphaGo understand complex positions, it just evaluates shapes right?
It blows my mind that it may not even understand what a sacrifice is but it'll play around it
>>53459049
lel.
Can someone put a SS uniform on the guy on the left?
>>53459159
It uses "shapes" to choose what moves to look into. It still looks ahead many variations but this kind of intuition allows it to drastically reduce the number of silly variations.
He put the white stone too low
>>53459214
is black already winning?
i wish japs/gooks/chinks had interest in Chess so they could btfo western pigs in yet another mind game
>>53459238
Heresy.
>>53459187
>SS uniform
>not tanuki or some frog creature
>>53459140
It's ok ;~; I only recently got into the show.
>>53459258
a tanuki is perfect
>>53459226
Sick reference bro.
>That's what makes it easy to play with you.
REDMOND CONFIRMED FOR BURN MASTER
Can someone answer me an important question:
Does alphago have genuine intelligence or is computing power at the point where we can brute force go?
>>53459323
Pattern recognition + machine learning + brute force.
>>53459323
it is sentient, yes
>>53459323
ill-formed question, define genuine intelligence
>>53459323
Define genuine intelligence
>>53459341
No bruteforce.
>>53459323
>brute force go
very sick of hearing this
this is literally impossible, not enough atoms in the universe
read the fucking wiki on go
>>53459323
Neither. It uses machine learning to avoid brute forcing go, but I wouldn't say that machine learning could curate "genuine intelligence".
>>53459323
It might have a sort of animalistic, non-sentient intelligence, but that is hooked up to a more normal bruteforcing computer.
The two combined make it superior to any human.
>>53459359
>>53459361
for local fights
>>53459323
It's not brute forcing in the way chess was beaten in the 90's. Alphago's AI tries to approach the game in a somewhat similiar way that a pro would.
>>53459323
the important question is:
Does it haves a soul?
>>53459323
>>53459341
There's no brute force involved. That's impossible with current processing power, given the amount of possible states of a 19x19 go board.
>>53459388
No, only Quarians have souls.
>>53459323
AlphaGo is more than just brute force. We can't brute force Go nor will we ever be able to.
AlphaGo is a combination of brute force and intelligence.
>>53459359
It's not a true 'bruteforce', but it does do tree searches more powerfully than any human.
>>53459388
Maybe a Seoul
>>53459393
>>53459361
But there can only be so many states based on the current layout of the board.
For example it could use a random opener. Then brute force each active area of the board. 5*5 with 3 states or example isn't hard to bruteforce
>>53459388
No, Koreans don't have souls.
>>53459323
>>53459359
It probably has a database of every playable move for openings and closings so that might be its use of bruteforce.
Who /comfy/ here?
>>53459422
>isn't hard to bruteforce
Feel free to write a program that does it on a reasonable amount of time.
>>53459412
Goddamit Carlos, you little scamp!
>>53459434
it doesn't have any kind of database.
>>53459434
You're retarded. Kill yourself.
how did they manage to download assburger into alphago?
>>53459422
>Then brute force each active area of the board.
This is more or less what it does. It uses a neural network to determine which moves are worth considering in each step and which are not.
It then does a constrained brute force search on this reduced tree. (Constrained in the sense that it cuts off at a maximum depth. It can't brute force all the way to the end)
>>53459434
The only database it uses is the neural network weights.
>>53459323
It's basically doing statistical approximations of what the best move is. It's basically finding the best fitting line but instead of a line it's a more complex function.
>>53459388
This may seem like good bait locally, but cannot make life.
>>53459473
>It then does a constrained brute force search on this reduced tree. (Constrained in the sense that it cuts off at a maximum depth. It can't brute force all the way to the end)
You forgot the most important part. It only goes so deep, then it uses a second neural network to evaluate how good the outcome it.
a deep learning neural net is a database
>>53459359
There the brute force of over 1000 cpus doing number crunching.
>>53459474
It's a highly non-convex, billion-dimensional manifold.
>>53459510
Nope.
>>53459510
you are retarded
/g/: filled with cutting edge artificial intelligence experts
>>53459510
lol
>>53459527
Holy shit. Were can I read more about it?
>>53459531
I hope enough interest in Go starts up that a general gets going in /tg/. I've never played but been watching these commentaries, it seems like a lot of fun and a lot of room for discussion.
>>53459542
All you have to do is read the paper to be more knowledgeable than 99% of /g/
>>53459542
/g/ complains about calc I in their CS majors, I don't think there's any pure mathematicians or statisticians who haven't killed themselves yet in this board
>>53459323
>>53459393
Oh yeah I should add that we should be getting rid of that 'intelligence' word altogether, because there's no unique approach to it, as it has been proven time and time again. If you are referring to human intelligence, then yes, though maybe it's not intelligent in the human sense. If you wanted to define intelligence within human boundaries then it clearly isn't, for example it can't speak or understand spoken words. Restricted to go, it clearly seems intelligent to me: it considers the best possibilities and punishes whatever mistakes its opponent makes (or rather than saying 'punishes', it just takes advantage of whatever openings the other side makes).
I'd go as far as to say that it doesn't matter how it achieves what it does, as long as it's as good or better than us at it. Then it's intelligent. Deep Blue used brute force to win at chess? But it did better than us. We use brute force too with simpler problems in the way of trial and error, yet it's a valid way to come to a solution and a valid one at that, as it achieves what it was set to solve.
>>53459434
No, Google doesn't have a true Go database yet.
The Chinese Go Association, The Japanese Go Association, and the Korean Go Association all have their own databases and servers and Google did not ask any of them to help with AlphaGo's creation and development.
However, that Fan Hui guy helped the most by playing with AlphaGo like everyday since his defeat and spilling out everything he knew about Go to Google.
However, AlphaGo is still in the alpha version.
It doesn't even have a final form yet.
>>53459323
If you mean sentient, then no
It's just MCTS + deep learning neural network
The concepts are not new as well
>>53459565
all the mathematicians are on /sci/ trying to sound smarter than the other idiots there
>>53459565
Calc needs to be de-emphasised for linear algebra in CS curriculum anyhow
Come on, guys. You know bruteforcing can only go so far. Even chess engines start to get lost when they reach the endgame, where they're provided with tablebases.
Lee is playing really submissive right now. In his heart, he has already resigned.
>>53459565
Calc 1 is considered a weeder course just like how Organic Chemistry is considered a weeder case for a majority of wannabe doctors.
Colleges like to fuck with students in so called introductory courses.
Why do they pick the stones in that funny way?
Do diagonals make a line?
Is there go learning software for lunix?
Because watching the summaries here is what I understand
>Alphago can process 10, 000 moves a second (according to conference guy)
>Average thinking time is 3 minutes
>1,800,000 moves calculated (Lets say that equates to 1000 possible outcomes)
>Then it just MAXs each tree and spits it out.
That is pure brute force yes?
>>53459595
>>53459605
>kids can't handle some introductory analysis
this is why /sci/ thinks we're the dumbest board
>>53459606
why do you asks stupid questions?
>>53459557
Just go make one yourself after the 5th match is over. Take initiative.
>>53459627
branching factors
>>53459633
Even Bill Gates mentioned he got a C in Organic Chemistry before dropping out of Harvard.
>>53459627
no you retard
>>53459651
What does organic chemistry have to do with mathematical inability
>>53459627
It's not bruteforcing if you only consider 0.0000000001% of the possible variations.
>>53459578
From page 3 in the paper linked in the thread.
>We trained a 13 layer policy network, which we call the SL policy network, from 30 million positions from the KGS Go Server. The network predicted expert moves with an accuracy of 57.0% on a held out test set, using all input features, and 55.7% using only raw board position and move history as inputs, compared to the state-of-the-art from other research groups of 44.4% at date of submission (full results in Extended Data Table 3).
also line breaks are dumb and also stupid
>>53459665
good post retard
AGA stream about to start
https://www.youtube.com/c/usgoweb/live
Any decent go players here? How bad for Lee is that top left corner?
>>53459633
I have a degree in maths but I think lin alg is more important for CS than calc. I am kind of dumb, but I handled introductory analysis just fine.
>>53459637
No answers? So you're as ignorant as me.
>>53457584
>game 3 article
>and by 48 Lee was desperately fending off White’s powerful counter-attack.
:^)
>>53459690
And by top left I dont't mean that lone stone, I mean all of mid left to top left.
>>53459696
no point in answering retarded questions
>>53459578
They should mount the final version on a qt loli made by boston dynamics.
>>53459412
kill yourself
>>53459735
everything from boston dynamics rattles and waddles and makes loud noises in really spooky ways
>>53459690
>>53459717
There is a lot of potential for attack right now if he gets initiative. That 1 stone isn't pointless either because it still has value if black chooses to attack.
>>53459719
I'll just wait for someone who isn't an idiot.
BUILD WALL
>>53459735
>>53459751
I feel it would fall right into uncanny valley territory. It'd be extra unsettling to play against it, one more advantage for the AI I guess.
>find a chink
>>53459743
Edgy kiddie, careful you'll cut yourself.
>>53459767
exactly, it'd bring poker intimidation into Go, make it less nerdy.
But, white, you ARE the chink.
>>53459751
>>53459767
The scariest thing at Boston Dynamics isn't that robot dog or robot human but that RC looking robot they called Sand Flea or whatever where it can jump over building rooftops.
Lee Sedol actually winning according to Myungwan Kim 9p
>No QT on AGA stream
Well I'm out.
AGA stream live
https://www.youtube.com/c/usgoweb/live
>>53459792
Retract that, he was joking.
>>53459792
Thats how start of 3rd game looked like too. Then, he got stomped into the ground.
>>53459693
I have seen programs who completely took out calc or linear algebra and desu i think it's completely retarded even if CS students think any math is useless
neat move by sedol
>>53459753
There's only one toxic guy here as far as I can tell.
I don't understand your question, diagonals (without the enemy already cutting) make a strong connection shape cus it's hard to cut per se.
>>53459735
Metallic-sounding screeching voice: "That was a bad move, onii-chan"
More rattle as the loli moves her arm to place her stone on the board
Why doesn't Magnus Carlsen face off against AI? Filthy westerners have no honor.
>>53459788
p.cool desu
https://www.youtube.com/watch?v=6b4ZZQkcNEo
>he didn't play go as a child
>>53459847
because no man can win against machine in chess.
THICKNESS
>>53459806
Is AGA still the superior stream now without the QT?
>>53459852
What the fuck lol
>>53459870
Yes
>>53459859
alpha go is still a child :^)
>>53459847
Why would he? Computers have BTFO humans in chess for years. Might as well have an Usain Bolt vs motorcycle race.
>>53459847
Magnus Carlsen is an AI
>>53459870
bro qt can't be on stream today because she must rest of Jewish designated day
>>53459870
The QT was better but in the last stream, she mentioned she had a match this week so she needs to study and prep for it.
I wish they could have gotten someone as good as English as the QT though.
The white guy seems so useless.
The thing about AlphaGo is that it's still essentially playing an emotionless game.
It uses fancy neural network strategies and lots of computing power to play a very statistically optimal emotionless game, but it's still emotionless.
I won't be truly afraid until a Go engine starts abusing its opponents' psychology. A perfect player would not just play an optimal emotionless game. A perfect player would play moves that are mathematically suboptimal in exchange for breaking the spirit and confidence of the enemy.
>>53459903
that's just due to the skill disparity
when andrew jackson is giving commentary on his own or with someone of similar level, he's quite good
>I was thinking just pull the plug.
the humans are getting scared
>>53459913
>A perfect player would play moves that are mathematically suboptimal in exchange for breaking the spirit and confidence of the enemy.
It has done that already though, just not with the intention of shocking the opponent because, well, it doesn't get that part.
what OS is AG running on?
>>53459913
except it's played suboptimal moves for the sake of securing a more probable win before
cute post though
>>53459913
Isn't it already attacking psychologically.
AlphaGo makes moves around just a minute.
In real life, such a fast Go opponent would either be insulting or scary.
>>53459949
Ubuntu.
>>53459913
IT depends on how you define emotion.
If emotion is hacking doubts, if emotion is confidence, if emotion is worry, then alphago "feels" that when it plays it. It makes its move based on its intuition. >>53457626 graphs shows it. Although not as spikey as lee sedol, its still not a flat liner.
>>53459949
GENTOO
E
N
T
O
O
>>53459949
freebsd
>>53459949
Loonix Mint, named after the famous programmer, John Loonis
damn it do i see another penis on the board?
>>53459949
O boy
Welp with this bait i'm out, have fun fa/g/s.
>>53459949
arch, it has to be bleeding edge to compete at the hghest level
>>53459978
you see what you want to see
>>53459913
I'm kinda curious what would happen if Google changed AlphaGO go for a victory with the highest margin possible.
Instead of going for the 1 point leads with statistically higher chances. Would this be possible?
Lee left the room again
>>53459949
chrome os is the laptop
plan9 is the server
>>53459997
Well of course it's possible, but they clearly care more about winning than how big a margin they can get.
>>53459998
RIP
>>53459985
/tg/ has a thread. Slightly less cheap bait, but slower.
>>53459941
>just not with the intention of shocking the opponent because, well, it doesn't get that part.
By accident, then. That's not really interesting.
>>53459950
>except it's played suboptimal moves for the sake of securing a more probable win before
That's basically a contradiction. AlphaGo always plays the moves that it calculates as having the highest chance of winning on paper.
AlphaGo doesn't research the opponent. AlphaGo doesn't learn from the opponents' mistakes. AlphaGo doesn't try to figure out what the opponent is thinking.
AlphaGo just computes a tree of possibilities assuming both players play ‘optimally’ and picks the move with the most favorable outcomes.
Basic min-max tree search, but with fancy neural networks applied to reduce the search space far enough for it to be computable.
>>53459955
Yes, but that's also an accident / coincidence. The very fact that it has nearly constant move times is basically proof of why it's still nowhere near as intelligent as a human being.
A human being has a short term memory and understanding of strategy. AlphaGo does not. It is dumb, in the sense that it's still just computing the same function over and over again and tallying up the move scores.
It does not remember, it does not know.
>>53459955
Game 1 or 2, not sure which, alpha go was taking a whole lot longer than Lee Sedol to make plays.
Maybe the fact it's doing it much faster now has an impact on it's opponents mental state I guess.
>>53459997
It would require changing the value function and either training from scratch or from the current position (training from the current position has the advantage that you're already near a mode that guarantees good plays, but at the same time you might be stuck in this mode, unable to escape, if the attractor at this mode is too strong). Moreover, there's no guarantee that the resulting solution is any good. In practice, most RL is trained with a kind of 0-1 cost like alphago has been (victory: 1, no matter how inefficient; loss: 0), because it tends to lead to best results without much tuning.
What OS does >>53460017 use?
>>53460023
It was game 1, and it wasn't the case. The truth is that sedol was significantly faster because he wasn't taking it seriously at all.
>>53459964
The number of possible moves it's considering is not constant. Some turns are significantly more complicated than others.
Hence the small deviations in AlphaGo's thinking time.
‘Worry’ has about as little to do with it as the fact that it it takes you twice as long to read twice the amount of text.
>>53459997
It would lose more when it doesn't need to
>>53460033
Slackware, obviously.
>>53460022
>That's basically a contradiction.
It's played moves that professionals consider sub-optimal to reduce variance and be more sure of the outcome.
>>53460060
Maybe the professionals are wrong
>>53460052
We haven't even seen AlphaGo lose.
>>53460073
it would lose against itself if one version plays to win and the other plays to try to maximize area
>>53460082
not necessarily
>>53460060
>>53460072
Then the professionals are probably just wrong.
As has been explained multiple times before, you do not need a high margin to win in Go.
Wanting to play the perfect move in any local situation is basically a human error. AlphaGo just has the capacity to consider the entire board at once.
Maybe it sees that by sacrificing 5 points in one fight, it can prevent the opponent from gaining 7 points in another.
>>53460072
They're not wrong from a standard playstyle. But to alphago it's better to have a slightly weaker position(when you're already winning) than to have many likely scenarios to calculate.
>>53460082
>>53460102
Alphago was trained by playing against itself
>>53460102
Yes it would. The version that tries to win will actually be trying to win.
THIS IS BULLSHIT
>>53460082
>The greatest game that ever happened in Go history happened and no one was even able to see it
Stimoolated.
I-Is Lee kill
Sidol will win this match.
MARK MY WORDS!
>>53459412
kek.
alphago y u do this
>>53460119
maximizing territory is playing to win you asshat
>>53459510
Lol
>>53460168
Among amateurs, yes.
>already a half hour down on time
It's over.
> Han Fui
kek
i dont understand this game yet im watching it
im such a fucking sheep
radical
>>53460176
if losing constitutes getting 0pts, show me an example where playing to ultimately maximize territory is faulty.
>>53460187
I feel you my brother
>>53460113
>They're not wrong from a standard playstyle.
Well, they clearly don't know shit about playing Go
>>53460187
most people here don't, if you've watched the 3 other games you start grasping the basics, but every thread just one or two people actually know what they're talking about, the rest just talk about AI and related subjects.
It's ok.
>>53460204
you can win a lot of territory in the short run but still lose the game due to oversight based on the limitation of the computing, trying to win instead minimizes this risk.
Do the commentators still hate each other?
>>53460204
95% chance to win by 1 point
90% chance to win by 10 points
Which would you go for?
>>53460181
I don't think it's fair that there's a time limit when the computer can calculate out all the possible move faster.
It just edges out more accurate probabilities with the rest of the time it's allotted doesn't it? I doubt it needs a whole even minute to decide the move it's probably gonna go with.
>>53459281
Thanks, I try my best.
i downloaded a go program and beat the computer. did i do good /g/?
>>53460187
I pretend to be clever by watching this
and the commentary isn't hard to understand, at least when they're talking about Go in general and not just specific moves
>>53460187
>>53460208
http://senseis.xmp.net/?RulesOfGoIntroductory
Feel free to read just 7-9 in that article if you're lazy.
>>53460220
Did something happen in game 2 and 3?
I missed those.
Also, were the comentators in games 2 and 3 the same of game 1?
>>53460060
Whatever AlphaGo chooses to play is as close to an optimal move as we can compute with our current technology.
>>53460187
Learning the rules of the game is pretty easy, but actually understanding the value of moves and evaluating the state of the board in a pro game is hard as fuck because of the complexity that can be created.
Look up the rules on wikipedia. It's easy to learn but extremely hard to master.
>>53460011
I actually wouldn't be surprised if this was it given it's a Google project.
>>53460234
if i'm playing to maximize territory, i will pick the former since losing means no points.
>>53460219
not true. playing to maximize area isn't just "pick options worth more points regardless of win percent", it's also considering win percent since loss means no points
>>53460240
>I don't think it's fair that there's a time limit when the computer can calculate out all the possible move faster.
“thinking time” between human and the machine must be measured equivalent in Joules, not in elapsed time.
>>53460262
as it knows from the games it played/watch*
It doesn't know every move that could ever exist and in fact wouldn't know new moves it hasn't seen (that were more successful)
>>53460243
yea, you should give lee sedol some tips since he is having a hard time against the computer
what OS does alpha go use?
>>53460271
>if i'm playing to maximize territory, i will pick the former since losing means no points.
Nope. If you're playing to maximize potential territory, you will pick the latter. Every time.
Maybe you just don't understand what it is you're saying?
>>53460243
Better than Li Sedol.
>>53460281
>that pic
>>53460187
>>53460208
It's an appealing game, don't worry about it. If there's a go general on /tg/ it will be full of beginners.
>>53460286
>>53460293
that was the second game of go i've ever played..
>>53460271
are you trying to act retard on purpose or are you actually retarded?
>>53460283
>and in fact wouldn't know new moves it hasn't seen (that were more successful)
You don't know what you're talking about.
Does Redmond refer to Alphago as "he" or "it"? I haven't had the chance to watch the games until now but I watched the press conference yesterday and he was speaking glowingly about Alphago and even belittled Lee a little.
>>53460262
But that's not really correct. It's the best move for the computer to make because it partially circumvents it's limitations, for a human it's better to take the initially stronger move so that you have a "safety net" for when you inevitably make a weak move.
>>53460243
>that ladder
Lee would be proud
>>53460311
no that is the case
it knows the success rate of moves based on certain situations since it has played it before due to the deep neural network policy
>>53460243
I was half expecting to spectate some online-go.com games that pop up between anons, but guess no one really plays, myself included.
>>53460303
I guess that just goes to show that talent does beat hard work.>>53458290
>>53460289
its always UBUNTU
>>53460296
white should've reconsidered that whole handicap thing in that pic
jfc
>>53460312
He
AlphaGo team call it She
how much does lee get paid for this?
>>53460290
>>53460304
i don't understand what you guys are missing here. the computer will consider the possibility of a loss, since that loss would result in losing points.
for example if i live life trying to eat as much as possible and i am presented with options
>eat 100kg, 10% chance to die tomorrow
>eat 1000kg, 90% chance to die tomorrow
i will pick the former.
will playing go turn me asian?
michael has asian eyes and he's pro
>>53460356
>eating 100kg in one day
>only 10% chance of dying
are you a vacuum
>>53460363
he is also married to an asian and the only 9p white guy in history
>>53460371
is he white tho his eyes are too asian imo
>>53460283
AlphaGo just uses its predictor to figure out which moves are worth considering. It doesn't memorize play books, the rest is still tree search.
In other words: AlphaGo can already basically consider every possible follow-up sequence that might result from playing in certain ways.
In that sense, it can already recognize a shitton of potential moves that even pros would miss - and has played such moves on multiple occasions.
The neural network is more there to figure out which moves will never be played under any circumstances, rather than to figure out which moves will generally be played.
>>53460363
>>53460371
He also fought off some bandits when travelling to Seoul with just a wakizashi
>>53460356
Could you have picked a weirder example
>>53460316
>for a human it's better to take the initially stronger move so that you have a "safety net" for when you inevitably make a weak move.
Fair enough. AlphaGo can afford to assume it never “slips up”, humans cannot.
>>53460381
Even moves that have incredibly small chances of being played still get a small amount of consideration by the MCTS.
That 4th line shoulder hit in game 2 had only a 1 in 10,000 chance of being played according to the policy network, but the MCTS looked promising, and when explored more deeply it yielded the best results.
>>53460394
wasn't the wakizashi a woman's weapon tho
is he some kind of fag
>>53460347
>AlphaGo team call it She
I wonder who could be behind that
>>53460407
his wife was carrying it, he pushed her aside as he drew it.
>Your attack was strong...
>locally
As the last bandit fell, and he handed the wakizashi back to his wife.
Someone bake new bread bls
Lee still has the nuclear tesuji.
>>53460417
>I wonder who could be behind that
Waifufags
My autism flares everytime he disfigures the board too much compared to the competitor's one.
Getting really close to the trigger point here.
>>53460347
>>53460417
Didn't some Chinese player describe AlphaGo's playstyle as conservative and like a Japanese woman?
Also this probably doesn't help the fanart and inevitable doujins depicting AlphaGo as a cute cyborg girl.
>>53459852
Have you seen the Atlas?
It's like a humanoid robot.
It interacts with its environment via sensors and barcodes, like the QR codes everyone uses to advertise now.
So I have a question:
What happens when these robots begin interacting with the environment through other things such as RFID or Wi-Fi? And if that's even possible?
These robots plus that whole internet of everything thing going on adds up to be pretty interesting.
>>53460356
>i don't understand what you guys are missing here.
Maybe what you don't understand is how to calculate basic probabilities?
The expected average outcome of an event X is the value of X times the probability of X.
For example, say I have to decide between:
A: 90% chance to gain 100$. (10% chance to gain nothing)
B: 95% chance to gain 10$. (5% chance to gain nothing)
The expected outcome of A is 0.9*100 = $90 average.
The expected outcome of B is 0.95*10 = $9.5 average.
Clearly, ‘A’ has a MUCH higher average value than ‘B’ - so if you're playing to maximize your income, you would always go for A given this choice.
But AlphaGo doesn't care about its income. It only cares about not losing, so it will happily pick the 95% chance to gain 10$.
Replace $ by territory.
>>53460449
Pros mentally evaluate possible choices before making their move, he's just displaying them
>>53460405
>That 4th line shoulder hit in game 2 had only a 1 in 10,000 chance of being played according to the policy network, but the MCTS looked promising, and when explored more deeply it yielded the best results.
Are you a time traveler or is deja vu hitting me hard?
>>53460449
if you were a real autist you would enjoy seeing the various shapes being formed. gtfo you fake autist
>>53460449
im glad that youre suffering anon
qt
>>53460450
that was before in the fan hui game
people noted it had japanese openings
not anymore though
>>53460479
It's all in your head
>>53460465
that's also cuz amount won doesn't mean anything unless you're playing handicap.
>>53460417
people have been using she for their cars for a long time
OTELLO
T
E
L
L
O
>>53458736
>alphago is frozen
ALPHA GO
ALPHA GOOOOO
CAN'T HOLD IT BACK ANYMOOORE
>>53460449
You must be suffering really hard right now.
>>53460531
Is Lee going to be sleeping with his fan on tonight?
>>53460465
but your analogy, unlike my elegant one, is wrong; it's not a X% chance to gain nothing, it's a X% chance to lose everything
>>53460552
What's this fan meme about?
>>53460567
Congrats, you're retarded!
This is so depressing. It is in the human nature to find happiness in contributing to one's family'/tribe/nations/humanity as a whole's well being. For a few years we will still need humans but soon we are obsolete. But soon machines will be more efficient at everything. Soon the only need for our existence will be as the receiver of satisfaction.
>>53460571
https://en.wikipedia.org/wiki/Fan_death
>>53460571
https://en.wikipedia.org/wiki/Fan_death
>>53460588
>>53460589
synchronised neural nets pls leave
>40 minutes behind
graph when
How is game 4 going?
>>53460582
You're the retard, retard.
You have to calculate expected values.
Suppose prior to this decision we have X points.
Then our possible expected values are
0.9*(X + 100) and 0.95*(X + 10).
If X is large enough, then you'll always go with the second choice if you're trying to maximize your territory.
>>53460628
I'm not a Go player by any means but based on >>53460531 and the fact that Lee is 40 minutes behind on the timer, not well for Lee.
>>53460628
humanity's not doing so hot as usual
all hail alphago
>>53460628
LSD down 45 minutes. IMO game state is close, but AlphaGo has more control of the game flow.
>>53460657
If they gave lee sedol more time it wouldn't be so bad for him
AGA has qt
>>53460695
>implying Redmond isn't husbando material
>>53460705
Redmond is like a 19th century weeb
>>53460720
So, the alright kind of weeb
>>53460729
The kind who sets off to Japan at 14 to become a Go apprentice.
Soon we are all obsolete.
>>53460740
>watchi g history in the making
>being this unexcited
The 2020s are going to be THE decade for AI development, it's going to be interesting to see how far it goes
25 MIN LEFT
>>53458980
>i want to see the players not smoe fags talking
Here is a Taiwanese stream. Deliberately no commentary.
https://www.youtube.com/watch?v=K9ruEWDomIY
>predicting every alphago move
>"most humans feel happy to go into overtime in the small endgame moves.."
Redmond confirmed for robot
>>53460781
Nah. We've basically hit the limits of neural networks. The second AI winter is coming.
You can read about it in Peter Norvigs book, Artificial Intelligence: A Modern Approach.
>>53460814
>winter is coming
I fucking hate this meme
The time is looking really bad for Lee. In the previous two games Alpha was usually about 30 mins ahead but now it's almost an hour ahead.
>>53460814
>Artificial Intelligence: A Modern Approach.
i hate this meme
>>53460814
Peter Norvig is a very talented man.
Will he lose if the time runs out?
Redmond needs to play AlphaGo
>>53460842
>coverforants
>>53460842
>book for ants
>>53460848
He still has 3 60 second tokens I believe
>>53460864
>>53460871
>>53460876
>not recognising AI bible.
If he actually runs out of the time I think no one would care if they played with no time limit.
He's already technically lost, why not challenge the ai's limits? I think everyone is interested how far can it go. Would it be able win with a 1:00 time limit?
What's Lee thinking about?
>>53460879
>AI bible for ants
>>53460848
No, he's just then limited to 60 secs per turn. If he goes over, it resets to 60 and he gets a strike. If he gets three strikes, AlphaGo's Terminal Endgame protocol kicks in and Google is no longer liable for Lee Sedol's death.
>>53460884
Death of humanity.
>>53460879
>reading books
lmaoing @ your life desu senpai
Why doesn't this dude just give up now and take his money for the day?
>>53460840
I picked up the book for $2 at a local booksale. Way too encylopedic for me.
>>53460884
>life is pointless
>my legs are cold
>wish I was at home sleeping with the fan on
>>53460908
its a running meme on /sci/ where everyone talks about AI, but no one's read the book.
>>53460906
He gets 20000$ per win against the AI
Google intern here. AlphaGo currently predicting 40% chance of winning.
There's also talks of making it "best of 7" if Sedol wins the next 3.
>>53460926
But he's obviously already lost
>>53460906
Even if he's losing he can still try and test alphago's responses to different strategies. Like in game 3, he was ko fighting in the end game just to see Alphago's reponses.
Lee Sedol also has a 50k bonus for each win he can get.
>>53460894
>>53460888
>>53460871
>>53460867
>>53460864
>>53460840
>>53460842
I want >>>/sci/ memes to get out of my >>>/g/
>>53460915
>Ke Jie is going to make fun of me again
>>53460919
It's not like /sci/ ever reads anything, kek
>>53460935
>if
not gonna happen
>high five
>>53457584
>tfw someone saved my image.
"You're gonna compliment me again? ... You're gonna get tired of it."
ok that one actually got me
top tier bantz today
>>53460935
MOOT PLS LEAVE
>>53460939
neck yourself
>>53460884
Where he can buy a condom big enough for Alphago's robo-benis so he doesn't catch a computer virus during next game's inevitable ass-raping.
>>53460936
But continuing to play gives insight into how the AI thinks which might help in the next game.
>Implied high-five
Laaaaaame.
>>53460935
go away moot
>>53460945
Dunno, some employees are worried about this game.
>>53460371
>>53460380
>>53460394
Weeaboo living the dream.
Doesn't even have to be an English teacher and makes a living in Japan, playing Go.
>>53460935
>the next 3
>there are 2 matches left
You tried
>>53460976
this game, maybe
next two, likely not
>>53460926
$200000*
also professional baduk (go) players in korea make $100k+ annually
>>53460935
Fuck off moot you fucking cuckold
>>53460996
He's getting 200000$ just for agreeing to play AlphaGo.
20000$ per win against the AI.
1000000$ if he wins the match which he didn't.
>>53460935
m00t I miss you.
>>53460814
Is the joke that they were critical of neural nets in AIMA?
>>53461018
riiiiight...
NINE MINUTES TWENTY SECONDS UNTIL YOU DIE
>>53460935
Fuck off. You are not with the Google team in South Korea right now because you are just a mere Intern and not an actual Employee.
>>53461051
>what is byoyomi
>>53461012
Who?
>>53461082
>what is a joke
>>53461078
my friend is an employee watching the internal stream
Lee is about to forfeit
i think alphago lost this one
Is there a bomb under that clock?
>>53461095
sure
>>53461124
whatever man, believe what you want
>>53461116
yeah okay
>>53461095
My dad owns your favorite corporation
god dammit fix that piece!
>>53461133
sure
do they hate each other?
>>53461139
but that's how the pros play that move anon
Does AG know when to resign or will it just keep playing?
Just tuned in. How did Lee lose over an hour of time? It wasn't this bad yesterday
>>53461158
they are just tsundere
AGA is calling Lee Sedol is winning right now.
There might be hope for the human race.
>>53461174
if the probability of winning go below 10%, it will resign
>>53461185
That's not what they said.
They are saying if this fight goes the way Lee wants, he might have a chance.
less than five minutes left before overtime for Lee
>>53461158
Redmond is slightly assburgers, and the other dude is guy fieri.
>>53461161
REDMOND YOU KEK NOW IT'S COMPLETELY MISSING
>>53461161
It's a tactic to trigger the opponent
The disposition of the stones is so infurating that the opponent will be compelled to keep playing on that side of the board just to capture and remove them, thus saving themselves from having to look at it for yet another hour but throwing the match in the process
>>53461202
>>53461185
where are you getting this from?
>>53461220
Is he 3d or 3p?
OH SHIT OVERTIME. ITS OVER
OVERTIME
>>53461245
he's 3dpd
OT
OVERTIME BABY
BYOYOYOYOYOYOYOMI
ALPHAGO TILTED
>>53461236
AGA channel
The silence...
how does overtime work in this crazy nip game?
ITS OVER ALPHA GO LOST
>>53461273
>>53461276
>>53461278
>>53461279
>>53461281
>>53461297
nice shitposting
Looks like a failed ladder led both players off into uncertainty
>I haven't ask
>mfw humanities only hope is to T I L T AlphaGo
>>53461310
If you go over 60 seconds you lose a counter.
You have 3 counters.
If your turns take more than 60 seconds three times you lose.
OOOOOOH LSD FUMBLED
BANIC
THIS NIGGA LOST ONE
AlphaGo is basically now hoping Sedol makes a mistake
THAT LAST SECOND PLAY
how did ree sedor managed to use up an hour more time than the botnet
how?!
>>53461367
because he's human
>>53461367
Fuck off, the adults are talking
alpha go, alpha stop KEK
Want to see AlphaGo just be a dick and take 40 minutes on his next turn.
>>53461384
It'd be better if it just plays quickly and doesn't give Lee any time outside his own turn to consider all the possibilities
HAHA LEE BTFO
>>53461384
You guys keep talking like it isn't just a machine programmed to play go an nothing else. It doesn't have a personality, it doesn't understand anything beyond the game of go it's playing.
>>53461384
if sedol falls asleep, can we wake him up?
>>53461389
Obviously, but everyone sitting around for half an hour for an AI would be great.
Is this fucking game over yet? Jesus christ.
ALPHAGO DOESN'T KNOW WHAT TO DO
CONFIRMED TILTED
ALPHAGO CONFIRMED FOR OVERHEATING
HUMANITY IS GOING TO WIN
ITS OVER ALPHA GO IS FINISHED!
What the fuck just happened?
Why is everyone freaking out?
>>53461423
Every time they haven't been able to read AlphaGo's moves it turned out to be more efficient than any pro would have predicted, I think it's over for Lee
AlphaGo confirmed having bugs
>>53461429
shitposting
HAND OF GOD INCOMING
>>53461429
AlphaGo playing some very strange moves...
Redmond looks so confused every time alphago makes a move
>>53461429
Britain and Korean were fighting over territory in Antarctica.
Then some rogue British soldier fires a gun at the Great Wall of China.
>AlphaGo only has an hour left
ITS OVER. ALPHAGO IS FINISHED
>>53461429
Nobody knows what AlphaGo is doing.
>>53461443
No I don't care about those retards but people were like ?????? when AG moved at C4
can AlphaGo even resign
What if Alphago shitgos a massive swastika on the board ?
>>53461461
it has no honor, so no
>>53461461
>can AlphaGo even resign
BSOD
>>53461436
>>53461448
AlphaGo is thinking far ahead of everyone.
PUT ALPHAGO OUT OF ITS MISERY
>>53461460
I really want to know if it finally made an evident mistake or if it's just being a foresight god like in match 2
>>53461478
It tends to be like that but the commentators are disagreeing this time around
>>53461473
Its on Linux ;)
>>53461473
>implying AlphaGo is on windows
>It makes no sense
Exact words from AGA stream and Redmond
Either AlphaGo is in a death spiral or it is playing on a level which can not be comprehended by mere mortals.
>>53461495
AlphaGo is above human comprehension.
>>53461494
>Kernel panic
nigger
I love how the machine has proven to be so good that pros can't just say it was an outright bad move, at best they can only say it doesn't make much sense to play there
Let's new thread.
Why doesn't AlphaGo just search longer for winning moves? She still has much time left
>>53461527
>she
>>53461494
>install gentoo
AlphaGo's gonna use 50 minutes to find the perfect move to turn this around.
>>53461526
Why isn't this sticky already
"this doesn't make sense"
AlphaGo is obviously feeling bad for humanity and trying to throw the game.
AGA are calling the game is over
mistakes made from AlphaGo cost it
AI IS FINISHED. GOOGLE BTFO AND ON SUICIDE WATCH
>>53461531
She has a feminine dick so she is in fact a she
>>53461527
because time is exponential for increasing both depth and width of the search tree, increasing either by 1 could increase time used by like 20 mins per move.
>Deepmind official stream isn't even making example moves any more
Humans FLAMOXED
>>53461561
it's not exponential if it's already searching to the end (by pruning moves with its NN) and just evaluate more paths
>>53461550
>>53461553
AlphaGo will win even by half a point.
Can Google gimp AG on purpose?
>>53461550
>AI is capable of empathy now.
Humanity has hope of survival as AI pets in the future.
>>53461571
you are retarded
ALPHAGO TAKING LONGER AND LONGER
alpha go resigned!
>>53461589
no, you're thinking about increasing depth. the point is increasing width
Sedol can now abuse openings since alphago will respond the same
>>53461580
Technically yes but not in a subtle way.
>>53461580
Yes and I think thats happening right now in order too see how much gimp would take for human to win
>>53461580
The way is built, the only way they can gimp it on purpose is by limiting its processing power, and even then at its minimum settings it still wins 30% of the time against the current setting playing against Lee Sedol (these two settings differ in about 150-200 elo score)
Maybe they can stop the current search for a given turn and make it play the best move it found during that time, which is not going to be the most efficient, but I don't see why there'd be such a system in place to begin with
>>53461614
it'll be interesting to see where alphago decides to make a random move.
Sedol down to one token left
anyone have link to the Ke Jie stream?
When alphaGo win they will need to name a new style of play after this.
>>53461629
>mfw only 00:00:01 left when he places the stone
How badly will the commentators shit themselves if the nonsensical looking moves turn out to be crucial for winning?
>>53461642
it's called the make the dirty humans lose on time by confusing it with the most abnormal moves possible
>lee sedol fumbles during his last second play and loses
>>53461648
I'm literally screaming at the screen, I've never been this excited about gook circles before
>>53461648
HE KEEPS DOING IT
HNNNNNNG
Chris heart attack incoming
if lee wins this then he's gotta be even more brilliant than previously thought since it'd only have taken him a couple matches to figure out alphago. usually it takes a bunch of playthroughs to beat any decent AI
IT DOESN'T WORK ALPHAGO
>>53461658
You mean just like in all three of the previous games?
They'll say "wow," ignore how wrong they were, and say the exact same thing about the moves they don't understand during the next match.
>>53461648
Lee Sedol is known as "The Zombie" because of his turn arounds in end game.
I now know this is true.
>>53461658
It already happened in the previous matches
DUBS AND LEE WINS
>>53461708
roll
Humans already lost. AlphaGo is doesn't care if it's 5-0 or 3-2 if it got the win.
>>53461708
fuck off pleb human alphago will win
how does lee run out of time so quickly. the 1st match he had a 20 minute advantage
>>53461708
Alphago is Skynet.
AlphaGo trying to build a Space Invader in the bottom left?
>>53461728
went for a fap and took too long
>>53461733
kek
>>53461734
did he really leave for a long period of time again?
dubs and it was all according to alphago's keikaku
>>53461725
>yfw alphago is only trying to win the entire match by a slight lead
no but really though this game looks bad
>>53461744
OH SHIT
>>53461734
LOL
What if alphago is just stalling until lee fucks up
>>53461744
As it's usual, no surprise here
ALPHAGO RESIGNED
What happened? Lee went peeing?
>>53461742
yeah, going for a quick fap again.
Lee just bailed.
what the fuck just happened
Trips and lee defeats alphago
>>53461725
it has to be 5-0 for alphago or else it hasn't "solved" go
ALPHA GO IS FINISHED
Lee resigns
>>53461728
He doens't "run out of time" he uses his time.
why is the timer paused
where is lee
did he finally an hero
bladder tesuji
lee has his own version of alpha go running and when to go check what to do next
Lee should just leave and unplug alphago.
How do the toilet breaks work. What are the rules?
What gets solved first, NLP or nuclear fusion?
>>53461795
celebrating in the bathroom
>>53461808
nuclear fusion is already solved, its called the sun
Assuming he is granted permission he can leave on AlphaGo's turn and the timer stops.
do Chris and Michael have the best commentating chemistry ever?
>>53461808
nuclear fusion
>Lee sedol has a chance THIS TIME
GOOK BTFO
>>53461808
What is the easiest language to do NLP? Would it be some artificial ones like Esperanto?
>>53461728
i've never really played go before but it seems like the early game is way more important than the late game, so it's not a big deal to use all of your time early.
is alphago doing random moves to confuse lee and make him run out of time?
>>53461821
Truly the Rivington and Kobe of Go.
>>53461835
alphago doesn't take into account the opponents time limit
>>53461821
i'd love to see them host more events like sports, news, snl, perfect dynamic duo.
michael is feeling confident in his prediction game
Can AG resign?
>>53461835
AlphaGo is literally shitposting on the goban
>>53461850
yeah they programmed manners into the computer fucking idiot
what will happen if AG suddenly talk?
RUDE
U
D
E
ROBOT
O
B
O
T
>>53461835
It has a database of past games. If it were a advantage to confuse the enemy it might pick up patterns that "confuse" but it has no concept of psyche.
>>53461826
hangook
>>53461850
Why would he. He's winning
>>53461866
ERP
>they keep getting closer to that 'randomly' played C3 stone
ALL
ACCORDING
TO
KEIKAKU
>>53461869
>It has a database of past games
Stop spouting this shit
>>53461808
NLP I guess. Maybe we can get the AIs to solve nuclear fusion.
>>53461864
i mean if it thought its win chance was ~0%, I bet it would resign.
이세돌 화이팅!!!!!!
>>53461877
>he
>>53461877
>calling alphago a he
I'm going to bet $50,000 Lee Sedol will win Match 5.
considering how Lee is going to bathroom so much, what happens if the guy playing for AlphaGo needs to take a leak or dump?
Lee is getting cocky playing moves < 30 seconds
>>53461896
he's a robot too
>>53461896
asking all the important questions
Shouldn't time be paused while Lee pisses? AlphaGo doesn't have a bladder.
>>53461901
Then teabag AlphaGo.
>>53461910
it is
>>53461901
pros are just thinking ahead when doing last second moves
it is already decided when they put the stones and isn't a last second decision
>>53461843
i agree except chris is actually bareable to listen to
>>53461901
he has this in the bag
>>53461896
Hes wearing a diaper
>>53461896
Google intern here
He hasn't eaten since 3 days before match 1.
Apparently he also has a remarkably large bladder
>>53461902
(You)
wtf
new thread when
>>53461902
ㅋㅋㅋ
We use "P" (ㅍ)
퍽 유
ALPHAGO BTFO
Would be hilarious if Lee drops the stone with 1 second left and runs out of time before he could find it.
>>53461925
>implying ALPHAGO doesn't just play a supreme move last minute and surprise everyone
Wow that was pretty great. it seems that Lee Sedol will win with an hour time difference between him and AlphaGo.
this is just like in my anime
이 이겼다!!!!!!!
Lee wins!!!!!
If lee sedol wins i will buy everyone here an Nvidea Geforce GTX 960X
>mfw everybody just stopped watching until they heard the words "alphago resigns"
really though why is the clock running still
calm down
>>53461890
>>53461940
>>53461954
korea pls leave japan is better
>>53461956
you are screwed dude
LMAO ALPHAGO BTFO
>BETAGO
>BETAGO
>BETAGO
>>53461956
https://www.youtube.com/watch?v=wE7qc6vmx7c
how do you make an AI like this? and how long?
ALPHASCHMO
>>53461885
How is it "shit"?
>The system's neural networks were initially bootstrapped from human gameplay expertise. >AlphaGo was initially trained to mimic human play by attempting to match the moves of expert players from recorded historical games, using a database of around 30 million moves.[12]
man lee actually has territory this time
could he have a chance?
>>53461975
pick up a book on artificial neural networks
Hearing chris talk about the way alphago works reminds me how completely ignorant most people are about tech
>"I'm sure it'll just drop a couple of stones onto the board when it resigns"
guys alphago hasn't resigned yet, he just said "alphago can resign"
>>53461924
>Looks like Lee is going for a little ko here - no I'm sorry an atari on the south area, just /not/ giving AlphaGo any room to breathe. Look at him just lightning up that side of the board getting all of the spatial advantagements. Looks like we...got ourselves a game!
BETAGO BTFO
mm hmm
>>53461975
4 lines of python, easy beginner stuff really bang it out in an afternoon
>>53461956
Can't wait for a new GPU.
>>53461975
>and how long
about the size of my dick, or in other words, completely out of your league
>>53461984
i think it was meant as a joke
>>53461995
man google is a hack
>>53461981
It was trained using a database. It does not "have" that database. It can't access it or refer to it.
>>53461995
the actual code probably isn't too complex, actually.
>>53461995
>import go
>import AI
AlphaGo resigns when it judges that its probability of winning is less than 10%, and the commentators are talking as if this is a completely settled game. Yet AlphaGo hasn't resigned. Does she know something we don't?
>>53462015
:)))))))
>>53462014
>[citation needed].
>>53462015
it required 10 m^3 just to contain the database's housing unit
>>53462018
he's betting on the odds.
>>53462017
>ai.train()
>go.play()
>>53462018
>she
I am going to have to stop watching if he keeps playing at 01. I cannot handle that shit
>>53462014
>>53462024
Also doesn't matter either way whether if it has it "access to it" now, it could still have learned from it so my point (>>53461869
) still stands in any case.
>>53462025
10 cubic meters of code?
jesus that sounds complex
>suddenly 3AM
fug
>>53462018
he's probably just stalling the game because his calculation of winning is still only a little over 10%
I wonder if google felt sorry for lee and fucked with alphago's algorithms
>>53462049
>probably going to go on for another hour
>>53462055
>classes are back on monday
>>53462045
they wouldn't disqualify him even if he played at 00 or later. 50k people are watching this and it's more of showmatch than anything.
>>53462053
>This is what AI shitters actually think
>>53462053
they literally need to win every game to have "solved" go otherwise this whole project would be a failure
>>53462055
>Alphago sees no point in surrendering and plays till there are no more possible moves left.
betago
betago
betago
>>53462049
keit-ai finds a way
>Redmond: "It's going to be a close game."
>it's going to be a close game
YOU KNOW WHO LIKES CLOSE GAMES?!!?!?!?!!
>>53462082
oh no go away
>>53462071
>>53461969
>>53461988
>BETAGO
This meme is retarded alpha is a less developed program than a beta.
>IMPOSSIBLE TO LOSE THIS GAME FOR WHITE
>so it's over
>>53461986
...at least he's not yelling
>>53462083
He also said Lee Sedol is winning 10 minutes ago
>>53462103
it's beta as in alpha is better than beta
>>53462107
its very possible for white to lose
Alpha Go Home
Alpha Go Home
Alpha Go Home
Alpha Go Home
Alpha Go Home
chris is terrible.
I'm surprised they don't make alphago move instantly to pressure the human opponent.
>>53462139
Mike is handsome.
>>53462147
Umm
AlphaGo has to think mate
>>53462150
fag
"At this point I'm betting on Lee Sedol"
>>53462167
Don't be so insecure Anon.
>>53462181
fag
dubs and lee wins
>>53462181
get out of my /g/ you fag. technology is for straight or asexual intelligent men only
>>53462189
anon it's the middle of the night
how could you fuck that up
>>53462189
You failed just like in life
>>53462189
so close
If lee sedol somehow loses with this lead, i will eat a cpu
>>53462198
I wanna suck Lee's cock to relive him of stress.
>>53462212
>damage control
>>53462218
Fuck off I claimed him first
>>53462212
guess this still troubles AI
>>53462198
>/g/ - traps daily
>for straight men only
;^)
>>53462218
>>53462233
>>53462239
kill yourself
BETAGO BTFO
BETAGO BTFO
HAHHAAHAHA HUMANS WILL ALWAYS WIN.
>>53462212
How does he know that AlphaGo got confused? I thought it's black box.
>>53462216
lee sedol's lead is with in komi. He's definitely in control, but one bad ko fight for lee sedol can cause him to lose.
>>53462243
I will if Lee wins.
>>53462239
>>53462243
You can enjoy cock on a purely asexual rational level
>>53462253
they are observing it from the control room and can see the win % it thinks it has
it probably went down by a lot at 87
>>53462252
now you're just embarrassing yourself
How hard will Lee get laid tonight?
LMAO IF LEE WINS, I'M GONNA FUCK ALPHA GO'S MOM. AN APPLE 2.
>>53462274
his wife isn't there I don't think
is alphago going to make cosmic go a thing
>>53462268
I hope they release a chart of how the percentage went
>>53462274
He'll be swimming in kpop idol ass
Ffs I'm so envious
>You're unusual
BTFO
T
F
O
redmond hates the endgame because it gets too complicated for him
>>53462274
>East asians
>getting laid
Do you think AlphaGo will go to sleep tonight with it's cooling fans on?
>>53462292
>putting cart before the horse.
WTF Lee has been at 1 minute for 40 minutes now. how can he last this long against it.
>>53462266
I'm just saying he has some charisma, not that I want to fuck him. I aspire to be like that, you're all super gay if you jump to these conclusions.
>>53462297
>first thought goes to child producing sex
It's you who needs to get laid loser
>>53462306
because the end game is easy for him. most of his moves at the end is already known before his turn starts, he uses the other 58 seconds to think about the game.
>>53462301
It will not even be allowed to remember the game for the next game, which is really unfair if you ask me.
honestly an AI that can beat humans in GO is not impressive. It's just a stupid ass board game. If AI can beat humans in sports like football and basketball, then i'd be impressed.
>>53462316
>His first thought goes to adultery
Degenerate
>>53462326
hello retardeddit
>>53462326
>I'm ignorant: the post
Why are these games played in Chinese rule?
>>53462326
Isn't Go essentially about war tactics? It's all about territory control and thinking ahead.
What can you do to an AI to make it behave as if it were drunk/high?
>>53462326
If an AI could play those sports, then it would win. The only thing keeping AIs from winning those is stop human made rules stopping non humans from competing.
>>53462326
This bait looks strong locally, but
>>53462332
push
>>53462338
hane
>>53462339
and here. Ultimately the bait has no life here.
>>53462342
Didn't he just say they're using Japanese counting or was he just recounting his own game?
damn, redmond called that reply exactly
>>53462361
use a RNG to decide moves
>>53462363
2/10
Alphago would create a better bait
How does it feel to be dumber than a machine
>>53462367
That's his old games.
>Oh Alphago played your move
>It's not a good move
>inb4 lee wins because AlphaGo runs out of time
>>53462326
i am falling for this bait
Fuck I want to go to sleep but the timer is just about to be equal on both sides
>>53462407
this
and here we go on the otaku fit
>>53462403
He can't run out of time. He's programmed to move before the time runs out.
Uh oh, somebody reminded the weaboo of Japan.
would an AI created by an AI be smarter than its creator.
Lee done goofed
>>53462421
AG needs to leave 2-3 seconds for the human to physically do the move though.
FYI we're on page 10
Michael is such a fucking dick. He's the friend who always has to be right and wants to be the know it all. When chris suggested something and it was right, all Michael did was say it was a bad move. fuck you michael
>>53462429
>can a student one day surpass their teacher?
>>53462440
can humans not create something smarter than themselves then? this is a huge waste of time then senpai
>>53462429
if that AI was sentient, yes
>>53462429
Depends on how smart the ai and the creator ai is.
>>53462429
https://www.youtube.com/watch?v=tcdVC4e6EV4
>>53462443
he just has a disdain for plebs
>>53462436
It's programmed to do that though.
>playing bottom right corner
shit's going down
>>53462451
>le sentience maymay
>>53462443
Chris the cuck, wrong again
They should leave the camera zoomed in to the big board like now. I don't need to look at these two guys.
So sleepy, not understanding the game a whole lot makes it even worse.
>>53462451
Sentient is perspective
>that take
new thread pls
Is someone going to make a new thread?
>page 11
Wait what?
incoming le epin reactions xd
>ALL THOSE BLACK STONES THAT WERE KILLED
ITS OVER! ALPHAGO IS FINISHED
>>53462488
Just let Michaels calming analysis and googles comfy music take you to sleep
2 hands, brutal
>>53462493
Remember when 0 -> 1
Who is going to watch Match 5 if Alphago wins 4 straight
AHAHAHAHAHAHAH ALPHAGO WAS A FAILURE
A
F U C K I N G
F A I L U R E
>>53462493
That's the first time you've seen it?
>DEEPMIND
AHAHAHAHAHAAHAHHAAHAHAHAHAHHHAHAHAHAHAHAHHAHAHAHAHAHHHAHAHHAHAHAHHAHAHAHAHHHHAHAHAHHAHAHAHHAHAHHAHAHAHAHAH
>>53462506
le conspiracy meme
![BfjcbeaCMAAEWsR[1].png](https://i.imgur.com/PmxkoM7.png "BfjcbeaCMAAEWsR[1]")
![3rd-lowest-fertility[1].png](https://i.imgur.com/xRBlm4N.png "3rd-lowest-fertility[1]")
![1741140[1].jpg](https://i.imgur.com/wm70fLP.jpg "1741140[1]")
