Can a machine ever truly beat a human? The best of machines vs

shameless self bump

It's this a post about chess, or just a general logic thing? Because a well designed machine will always beat people at a game that can be "figured out"

>>716739201
False. A human will always be able to think outside the box and go beyond any training. Computers can only do what they are told to do. A computer is limited by whoever programs it.

>>716739304
If a game can be figured out then the computer will always win. Tic tac toe has a set number of possible outcomes, and a program that knows these will always make the optimal move. Take a math class

>>716739304

Then please explain how computers calculate with high numbers infinitely faster than humans?

>>716739480
Tic Tac Toe is always a draw, that is a shit example. A game of Chess is almost infinite. There are more possible move combinations in a 40 move game of Chess than there are particles in the known universe. A Chess Grandmaster will always be able to out think a computer as the computer is limited by whoever programs it.

Theres already been a computer That has beat a grand master in the ancient Chinese game 'GO' , Its supposed to be more difficult than chess. So game wise, computers have already beat man at one of the arguably toughest games to play

>>716739645
Calculating high numbers is not a game of Chess. It's just number crunching.

>>716739645

The burden of proof is on you, faggot. Google deep blue or even that one dumbass Jeopardy machine.

>>716739785
Source?

>>716739836
That Deep Blue bullshit was a hoax -- it was being controlled by several Grandmasters. Check out the movie Game Over: Kasparov And The Machine.

>>716739304
>what is AI
>what is SAI
Have you been here before 4chan fucks with the internet AI little girl?

>>716739874
https://www.scientificamerican.com/article/computer-beats-go-champion-for-first-time/

http://fortune.com/2016/03/12/googles-go-computer-vs-human/

Program was named AlphaGo, theres a youtube video of the match out there too

>>716739304
Literally take a math class

>>716739874
Thing's called AlphaGo - you go from here. It's not perfect, but it beat the human champ.

>>716739744
If its finite then its not beatable.

>>716739874
>>716740114
https://www.youtube.com/watch?v=mzpW10DPHeQ
Theres the final game, Yes its long as fuck but the results of the 5 games are at 5:32:04

>>716740114
Interesting.

>>716739824

So what. It is outsmarting humans in that way. Why wouldn't it be possible in the future that a machine could do other, or all things better than a human?

> It's not chess, so your argument about something else is invalid.

>>716740309
>>716740312
Personally i only watched part of the video just out of interest in the game and how its played

>>716740328
Machines are better at some things, a lot of things in fact, but anything that requires free thinking, a human will always win.

>>716738912
Yes, we are approaching the singularity at a fast pace and soon artificial intelligence and Transformers won't be Science Fiction anymore.

Beating SNES Streetfighter 2 on 7 star difficulty is proof that man beats machine.

Not OP but I sort of see the point he is trying to make, although Chess is a very bad example because it simply requires the calculation of every conceivable outcome to win, and a machine can calculate that faster than a human brain.

>>716738912
>>716739304
Well,if the machine is made it to think outside the box,maybe.

if it requires only logic, computer beated humans long years ago. if it require more creativity, then program is as good as its programmers.

>>716740473

You do realise that the human brain is made of components which guide electric currents.

Say in the future we could replicate a human brain with components and electric currents to guide this.
Why would it not be able to "free think" like humans do?
Our brains are made up of stuff.
Stuff can be replicated. Stuff can be improved.
It's inevitable that we will build AI that will surpass us.

>>716739744
>almost infinite
This is why a computer wins. A computer can access every possible permutation of the outcome of a chess game at any given state of the board. Unless the programming is flawed and the human knows how to exploit these flaws. The computer has a great advantage

>>716740924
Can a machine every truly think outside the box though? True AI does not exist yet. When true AI has been invented, the scenario of Skynet in Terminator or the programs in The Matrix are a very likely outcome. They would make humans extinct.

>>716741007
A great advantage but, in my opinion, a top Chess grandmaster would always be able to outwit the machine eventually.

>>716739744

Actually, nowadays AI can easily beat Grandmasters.

Pretty sure they could make true AI if they wanted to. The feds are keeping that shit under wraps.

>>716740473
Chess is not a free thinking game. There are clearly defined rules that must be adhered to. A computer can use these rules to always make the most optimum play

>>716741339

>>716741252
deep blue beated kasparov in 1997. Nowdays top grandmasters have no change in classical game as computer calculates way faster.

>>716741346
The unpredictability of a human brain would always be able to remain one step ahead though?

>>716741346
This
Eventually a sufficiently good computer will be made and humans can never win.

>>716741416
The first time they played, Kasparov actually won 4-2. The machine won in the rematch. If a computer is so infallable, it would just just win every single time.

>>716740581
Only because of forced flaws. Street fighter would not be fun if the computer would reflex parry every single one of your attacks and wait until it detected a punishable opening to destroy you.

>>716741104
As long as Keanu Reeves is still alive, we have a chance.

>>716741474
No because however unpredictable a computer brutforcing all possible moves will not "allow" a move to be unpredictable.

If it has bruteforced everything between every move.

>>716741346
in classical game both players have max something like 3 hours to play. Top chess software beats humans every day.

>>716741568
that was different version of deep blue. they updated it for 1997 game.

>>716739874
>where the fuck were you when it was all over the news???

Why are Russians so damn good at Chess? Like all the greatest players ever are mostly all Russian.

>>716741104
1) How do you know true AI does not exist?
Does the rabbit in a cage in the zoo know that it is in a cage in a zoo?
2) Steven Hawkin has said that he is 50% sure this is what reality is like right now.

>>716741474

Unpredicatable moves are not the same as the best moves.

The machine will calculate all moves that can be made (including the unpredicable one). And always make the best move possible.
A human is just not capable of that. We way worse at knowing all possiblities and then picking the best one.

>>716741932
1) The best is from Norway
2) cause they play that shit in school. Like PE, math etc etc. Its a mandatory subject.

>>716741932
that because chess is hugely supported by government. it was part of soviet era thing.

OP is talking horshit.
But.
A computer, true AI or otherwise, has to be built and programmed by a human, so if the computer wins, it is still a human achieving the victory as it was created by a human.

>>716741474
Unpredictability doesn't matter when a computer can see every possible outcome of your next moves until the game ends. Unpredictability actually is better against humans than computers because with humans you can get inside their head and make them second guess their strategy. A computer doesn't care. It simply makes the most optimum decision in any situation

The unpredictability of the human mind, and instinctual roots, will inevitably survive against any AI or programmed thing that does not have the ability of free thinking*

File: hqdefault.jpg (21KB, 480x360px) Image search: [Google]

21KB, 480x360px

Read "The last question"
Written by Asimov.
Then come back.

Its a short story. Made me look differently at (future) AI.

In the match where Deep Blue beat Kasparov, it took 15 minutes to make one single move at one point. Trying to fathom how many possible outcomes the computer calculated in that time is beyond the comprehension of any human mind. The computer absolutely crushed him in that particular game by the way.

>>716742081
and humans have to even put power cord to power socket! lazy machines.

>>716742374
>press off switch
>human wins again
>spin a 360 and moonwalk the fuck out of there

>>716738912
Mine can :)

It's pretty well-documented that just as computers can figure out a game, so too can a person figure out a computer - and relatively quickly. Unusual openers, surprise moves, etc. throw a computer off its game. Part of this has to do with the fact that programmers aren't chess grandmasters, which puts the computer at a disadvantage but not vice versa. Part of it is a consequence of how a computer is innately designed. That said: how advanced we talking? If you think a computer could reach levels of intelligence comparable to man's, and you couple that with its ability to see every possible move ahead, it becomes a tough situation for fleshies to win in. I dunno, not sure there could be a proper answer.

File: jkj.jpg (21KB, 551x223px) Image search: [Google]

21KB, 551x223px

>>716742461

>>716741474
NO, JESUS CHRIST GET IT THROUGH YOUR THICK SKULL
>unpredictability being a valuable trait in a game with fixed outcomes

>>716742544
Agreed. What OP is saying isn't true, but it isn't as clear cut as a computer just winning every single time. Where we are at at this point in time with computers, it's a pretty even match in my opinion. In the future, who knows.

>>716740954
The issue is it wouldn't have a conscious as we perceive it
We don't understand the human brain well enough yet to properly treat issues with it
>let alone build an electronic version

machines can't paint great art like vincent van gogh, or compose great music like mozart.

>>716741104
I don't believe true ai will ever exist because I don't think we'll ever fully understand our own brains

File: oFYMVaI.jpg (32KB, 500x376px) Image search: [Google]

32KB, 500x376px

>>716741932

Pretty sure the Deep Blue thing isn't all that it seems. Apparently some of the moves the computer made were very human like and it made mistakes. If it was a true computer, it wouldn't make a single mistake (as long as it is always programmed to go for the win).

>>716742745
Wrong.
https://en.wikipedia.org/wiki/Artificial_neural_network

>>716742042
Total BS
The second part, I:ve no idea who the current chess grossmeister is

>>716741104
>yet
you're answering your own question, stop looking like an edgy dumbass

>>716742838
Can you?

>>716741252
>implying chess is nothing more than a huge math problem
If the game actually required free thinking then yes
However what we perceive as free thinking is simply the best math solution we can come up with at the time

>>716741339
We don't know enough to properly make intelligence

>>716742946
that's the conspiracy theory around the 1997 game and i don't buy it. Anyway, it doesn't matter, coz chess softwares beats humans nowdays.

>>716743061
You greentexted wrong.
>implying chess is nothing more than a huge math problem
>implying that chess is anything more than a huge math problem
FTFY

>>716743002
It might be Azabajan.
But in x soviet contry they play chess like you yankees play handegg or baseball.

>>716743061
There is more to a game of Chess than just maths, otherwise the top math professors would all be top Chess players. The two worlds are pretty different. Chess requires different tactics, attack and defend, counter attacks, etc.

>>716742154
yea, only if by sheer luck the human chooses a unpredictable move that allows a win.
There are loads of unpredictable moves, always. And 99,99% are complete SHIT.

I believe the machines of the future are so 'logic' so, 'smart' that the 'unpredictable' move you are referring to, is very predictable. Since it is supposed to be the best move, it's only LOGIC that you would choose it. Then the machine already has thought of 4 different counter moves..

You lost.

Stockfish can beat any grandmaster in Chess.

>>716743207
Yes, it is just a conspiracy theory, but how can you explain a computer making mistakes in a game of Chess? It just wouldn't happen, unless the computer was being manipulated in real time by a human.

>>716743256
>There is more to a game of Chess than just maths
No.
There are too many board iterations for human beings to memorize.
>13^64 maximum possible positions, because each piece could either be black or white.
Did not verify number, copy and pasted.

>>716742745

I'm talking about future computers. There's no reason not to believe it's possible to build these in the (near) future.

File: hand2.png (82KB, 2416x1812px) Image search: [Google]

82KB, 2416x1812px

>>716743015
yes. (pic related)

>>716742838
but they can make art.
just becaue you dont think its beautifull.. doesn't mean its not good art.

They make better abstract art then humans imo

>>716743292
Predictability within unpredictability. An interesting concept.

>>716743482
Prove it

>>716741835
Which may or may not have been 100% legit
I think a grand master could tell the difference in his opponents moves
Especially if 97% of them are best case scenario moves and robotic in nature
>I don't think Kasp was just being salty

>>716743482
show me then what kind of art program makes.

>>716741932
>what is state sanctioned

When the first true AI will appear, it will probably ask if it is alive. But it would certainly ask itself what can end its life. The answer will of course be humans. That is why I think some sort of conflict is inevitable.

File: duh.gif (2MB, 385x204px) Image search: [Google]

2MB, 385x204px

No man on earth can compete with the Hitachi. /thread

>>716743225
Idk what the shitty satellite states are up to, but I know for a fact its not a mandatory thing, neither in russia now nor in the soviet union. It was pretty loved, granted. There were and still are (though nowhere as many) loads of chess clubs in schools and universities. And why not, it's a good game.
The modern chinese version is fun, too

>>716743676
The answer is also other machines, degradation of hardware, and random chance.
Humans can counter and rectify all three.
Our usefulness outweighs our potential conflict.

>>716743397
What humans interpret as a mistake is simply the computer calculating a move that looks bad on the surface but has a net positive turnout many turns ahead that a human wouldn't recognize

>>716742705
Couple what you've said with the wiggity boundaries - i.e., what if man uploads himself into a computer? what if man computerizes himself (shoutout to the guy who bumped The Last Question)? what if computers get advanced enough to qualify as, basically, number-crunching logically-oriented people? - then I don't know if this competition or this question is valid in the first place. I mean, if you want to know how things are right this second... look the answer up yourself. If you want to know the future, better place your guesses before the future hits us.

>>716743002
Magnus Carlsen
Not Russian

>>716743578
Agreed.
I think Kasparov was a very sore loser but I just can't accept that a computer would ever make a mistake if it was programmed to calculate the best possible outcome in every single position.

>>716741947
Because I don't think true ai would ever allow us to build ai
We simply don't understand what makes a brain enough to make ai
>implying Hawkins can't be wrong with 50/50 odds
He should really just toss a coin... sucks he can't do it himself

>>716743645
>>716743558

Ever heard of google?

>>716743790

Ouch right in the sack.

>>716743937
sure. im not sure what kind of art you are referring to.

>>716742080
I think it probably hails back to earlier than that but became a thing of the people as a symbolic thing

>>716743516
Machine's dont think in predictable and unpredictable.. just 0's and 1's. They will never be surprised, or taken by surprise. Just because that is not something you would program into a machine. It just 'is', it has machinery equipped to mimic our 5 (or 6 or more according to some..) senses. And constantly collects data.
Something unpredictable cannot happen. The machine just registers facts.

I guess the only way to 'surprise' a machine is to somehow 'hack' the senses, so that false data is gathered. Then the actions of the machine will me predictable, and you can take that to your advantage.

>>716744079
or what program that makes so great art as you claim.

>>716743397
It was being manipulated, but between the games, and thats not a conspiracy but a fact. And it's completely fine by the rules - learning effect and all that. Kasparow didn't want another rematch after the whole thing. ( Did at first whilst some were screaming fraud but changed his mind)

>>716743578
>grand master could tell the difference in his opponents moves
Implying that a computer picks moves in a way that a chess player can fathom. The methods that a computer would use to make its move is completely different to how a human thinks

>>716744093
but sometimes my windows surprise me with bluescreens and errors.

>>716744079
search: computer generated art, or anything similair. Personally, im a fan of symmetry. Computers are, in fact, very good at this

i was hating ai before it was cool to hate

>>716744149
>"AI Generated Art"
>http://lmgtfy.com/?q=AI+Generated+Art
You're a fucking retard.

>>716743909
What is a "mistake" though? You cannot possibly say for 100% certain that whatever move you are referring to was not a delibate action by the computer, in an attempt to outthink and outwit Kasparov.

>>716744093
The same can be said for humans, if you perfectly replicated a human mind with some massively parallel processer and fed it sensory data, we could perfectly predict how it would respond.

Humans are only unpredictable because we don't completely understand how the mind works yet.

>>716744149
>>716744079

https://www.google nl/search?q=artificial+intelligence+art&ie=utf-8&oe=utf-8&client=firefox-b-ab&gfe_rd=cr&ei=Kd5iWJa0Fq3c8Af27a-QBA

>>716741104
>Well,if the machine is made it to think outside the box,maybe.

>>716744302
Because you keep forgetting that viruses and other malware are being installed while watching porn and shit. If you constantly think of that fact, and the chance it gives to BSOD your machine.. you will never be surprised by that bluescreen.
A machine would remember.

>>716744245
humans know instantly what is considered as bad move. chess software has to calculate every possible option in every position. that's lot of work to humans to do. like 5 moves ahead.

Are we talking only about chess? Because I remember someone building a machine to score perfect 3 point shots in basketball, which began missing after a series of perfect shots. So machines are not perfect so far.

>>716741947
>implying that hawking isn't already a cybernetic brain running in a shell of a human

File: CT_logo2.png (14KB, 320x177px) Image search: [Google]

14KB, 320x177px

>>716738912
Yes of course, dude. Your question reveals a fundamental misunderstanding about the "machine" however, that I feel should be clarified to give you a better understanding of the situation.

when you play a computer, you aren't really playing a machine, you're playing a program made by a team of very experienced players and programmers who have done their best to have the computer memorize all the best moves and be able to formulate what patterns of moves to go for.

You're actually playing against the coordinated efforts of an organization of other humans whenever you play a computer game. It's not actually the machine thinking up all that shit. It's other people who have set it to do what they want it to do.

TL:DR You're just asking the wrong question. Machines don't really think for themselves, they are programmed with other people's thoughts.

>>716743000
Checked but that is literally just free calculating things
It generates it's own problems and isn't anywhere near what true ai is

FROM YOUR SOURCE:
The goal of the neural network is to solve problems in the same way that the human brain would, although several neural networks are much more abstract. Modern neural network projects typically work with a few thousand to a few million neural units and millions of connections,
> which is still several orders of magnitude less complex than the human brain and closer to the computing power of a worm.

New brain research often stimulates new patterns in neural networks. One new approach is using connections which span much further and link processing layers rather than always being localized to adjacent neurons. Other research being explored with the different types of signal over time that axons propagate which is more complex than simply on or off.

TL;DR
You're still wrong and actually gave me *citation needed for my argument

Good job kiddo

>>716742870
we only have to get to the point where we can design and program a computer that can better design and program a computer then we do. When that happens the computer will eventually figure our brain out

File: BN-MV785_DDream_G_20160301074855.jpg (69KB, 553x369px) Image search: [Google]

69KB, 553x369px

computer art is shit.

>>716743209
>implying I was trying to use green text to imply and not just to color my text and set it apart

>>716738912
> OP doesn't realize that machines have been beating humans at games of strategy and other challenges since at least the 80s.

Welcome to the 21st century, OP. You will forever be known as a Luddite.

>>716739304
> computers will only do what they are told
Someone isn't up to speed on AI learning.

Humans barely program our current AI systems. They basically code themselves.

>>716739744
A computer beat the top-ranked 'grandmaster' at chess in the 80s, dude. Where the fuck have you been?

Most people have an infantile understanding of computers and the internet, at best.

Computers don't do anything they are not explicitly instructed to do if they are operating normally. That means, it's not your computer coming up with those game winning moves, it's an actual human being who has instructed that machine to operate in a specific way.

Do you think Machines create art? No, they are created to make art, among other things. Machines don't have a will. They are simply tools that are manipulated. Sometimes like puppets, and you know how people love to pretend.

uuuuuuuuuuuuuuuuuu

>>716743209
Implying I'm not fucked up and shitposting on b

>>716743000
But is an artificial neural network able to make cognizant decisions? Can it recognize it's own decisions and change them before the decision is made?

Say for example, I am angry. Someone could ask me a question and I could retort rudely, or perhaps I think "Maybe I need to calm down" and decide to not respond rudely.

Can an artificial neural network recreate the cognition and empathy that sets us apart from other mammals? Can they go beyond instinctive thinking and actually ponder or think about something beyond established facts that can be absorbed?

I don't think we are anywhere near simulating that. I think it MAY be possible... but probably not in our lifetime.

>>716743209
>RUMadBro¿

File: cleverbot.png (7KB, 206x255px) Image search: [Google]

7KB, 206x255px

cleverbot do not surprise me.

>>716744386
I dont know, since machine's think in 0's and 1's, it's either thisORthat.. blackORwhite.. leftORright. It would calculate the best option according to the data it has, and 'chooses'.

Humans dont have to choose between black or white, we can go grey.. lots of shades of grey to choose from. That makes us better thinkers than machine's, for now.

There will be a point when the machine's see the grey scale, and more, since we can give them sensory equipment we cannot ever hope to have. Certain kinds of radiation, meta-data from surroundings, and ultra fast communication with other machine's.

We have our limits, the machine's we will built (and eventually, the machines built by machines) will surpass our calculating capabilities, physical power, speed, sensory capabilities and so on..

If humans don't fuck up, and destroy themselves, it would be so interesting to see what the world, the universe looks like in 4000 years. (If i have the intellect to understand what the fuck is happening there, ofcourse.)

>>716743256
>implying the human brain can handle the amount of calculations in regular time for each turn out of opening

>>716745087
but that's not true. only in 1997 computer beated human in classical game.

>>716743256
Top math professors are renowned for being able to solve complex problems requiring new solutions that can't be plugged in.

This means they are good at disassembling complex logic problems into simple pieces. It doesn't mean they can math their way through every "logic" situation that comes their way.

Like any competitive game, chess requires a great deal of empathy and foresight.

A chess player will move their piece in a way that protects their pieces.

A good chess player will move their piece in a way that it puts pressure on their opponents pieces while protecting their own.

A grand master will move their piece in a way that it will prompt a specific response from their opponent. They recognize their opponent's habits and quirks from observations made during the game and use this to determine their line of play and counter-act it accordingly. Likewise, their opponent will adapt to different strategies to counter.

This goes much beyond "math". Math can help you narrow down the optimal line of play but the human factor is very important.

File: puppyslugs-610x458.jpg (132KB, 610x458px) Image search: [Google]

132KB, 610x458px

>>716743937
i googled it but that was shit art. perhaps you and i have different taste.

>>716745260
Thousands of years of evolution and technology, and this is basically where we are at kek.

A reminder to the people talking about a singularity. People have been talking about a supernatural rapture from the gods for eons. This is nothing new, but it is complete fucking fantasy for now. No machine knows what it's doing. It's just like a gear being turned or series of switches being flipped. It doesn't "think" it simply does, and one of those things it's instructed to do is to appear as relatable and as human as possible because that's what people want to buy and that's what sells.

Machines don't create art, they don't love, they don't kill, they just are. It's their operators and programmers that share their life, spirit, and intellect with the machine and animate them to the point where people who don't know anything about computers can be impressed at how "smart" their phone is or how innovative some photo filter is or whatever the fuck.

>>716744245
That's my point
He could tell when a move was very human like (ie reactionary to a single move with little regard to the rest of the boand)

>>716745639
seems like it copies some image in internet and paste those balloons in it. very inspiring indeed!

>>716738912

Negative, Ray Kurzweil predicts that the singularity either in transcendence or in true AI will come within the next 20 years.

A shift in genetics will be inevitable thanks to existing social class differences. If you haven't noticed, the gap between the lower and upper class (lol at the non-existent middle class) is growing wider and only those with access to bleeding edge tech can succeed.

>inb4 but there is always a shift in jobs

Yeah, keep regurgitating that. For every 10K jobs that AI replaces, 1 new job will be available for humans.

>>716744203
They dismantled the machine after they declined his offer and denied to let him see anything
IBM was very shady about the whole thing

>>716745124
how do you know?

when does consciousness occur?
There is an egg and a sperm cell, they grow into an embryo which becomes a foetus and then gets born into an infant and then ages and becomes a child and a teenager and an adult who we consider fully conscious. The consciousness grows and is not there at a certain point. It goes through all the phases of which some will be comparable to the consciousness of an insect or a goldfish or a monkey or a computer program or other systems like the stock market or even /b/ as an entity

its already there you faglords, everything is alive and the highest consciousness is the universe as a whole aka god

>>716745764
But there is no way of proving that that wasn't a deliberate tactic programmed into the Chess computer to imitate.

>>716741104
all we need are efficient algorithms for framing data gathered from the AIs environment and heuristic algorithms for planning the best response. Many AIs exist today, many of them are exceptional at what they do and are even judged "creative".

The people that wrote their code have no idea how the AIs come to the conclusions that they do because programmers pretty much only program success/failure conditions into it and "train" the AI to an acceptable performance level.

File: the b.jpg (41KB, 530x402px) Image search: [Google]

41KB, 530x402px

>>716738912
I'm calling what you call a machine a computer.

A computer designed to play chess perfectly will beat a human. This is because, despite the fact that a human might be able to adapt to situations and look 'outside the box', there's only so much that can actually help while you're doing something that has defined rules like chess.

My vote is on the computer for beating a human at a game with rules. Computers are the best at figuring out how to best follow rules.

In something more ambigous though, like a war between humans and computers, I'd vote for the humans because there aren't any rules so thinking 'outside the box' becomes much more advantageous.

But I don't like to get all human-centric. We have a big ego as a species, but there's no reason to assume we can do anything. There's very possibly a limit to what humankind can accomplish.

>>716744802
To an extent yes but you could make a net to "freethink" a chess game but it would ultimately brute force everything

True AI will only exist when a machine feels the urge to fap.

>>716745898
perhaps they feared that kasparov would tell rival computer manufacturer IBM's secrets.

>>716739304
This. Like it or not.

Even the Deep Blue v. Kasparov stuff has been effectively debated. If you watch those matches, you'll see that Deep Blue suddenly changes its style in a totally counterintuitive way in the end game.

It is believed that either it was a glitch or (more likely) a combination of the computer's logic and modern chess wisdom of Grandmaster Joel Benjamin that beat Kasparov that day.

Could Deep Blue beat anyone in this thread? Yep. Could a chess computer beat Garry Kasparov in his prime, fair and square? Nope.

IBM simply had too much on the line to allow Kasparov to win.

>>716745614
This is why I think Kasp wasn't just being salty
He noticed when a move was human

>>716746490
but how would they reproduce?

Yes. Chess. No human can beat certain computers at chess. 10/10 times.

>>716745124
How do you explain the programs that people leave running for a few years focused on a task (chess), and over time gets better at it? The programmer of the original program wouldn't know how to beat an experienced player, but the program would be able to 100% of the time. You can't say a programmer programmed those moves in.

>>716746039
>implying you'd program a flaw in your program to beat a grandmaster
>implying it wouldn't be easier to change the coding during the match

>>716739744
Show me a man beating deep blue you fcuking retard

>>716746469
>freethink

What do you think you're talking about, dude? Computers don't think. They think as much as a car engine, or a motorcycle.

what exactly do you think "freethink" is?

>>716746549
kasparov was just mad coz he lose. Before that 1997 game kasparov said that computers never beat humans in classical chess game. He was just mad to be wrong, stubborn as he is.

this whole human vs machine thing is kind of stupid, because basically every machine is created by a human being and it's entire behaviour and thinking is based on ours, so every match of chess is a match of human vs human

>>716738912

Understanding and computation are two different things. A computer may be able to beat the best human at chess but it doesn't actually understand what it's doing. It's merely using algorithms to calculate the most ideal play based on the rules of the game. It doesn't understand the game, what it's doing or how it actually won. This is the critical difference.

>>716746392
Computers DO NOT have the ability to see every possible ending to every possible move, not yet anyway. There are more possible outcomes for a game of Chess than particles in the known universe, as another anon said. Computers have a great, immense, almost unimaginable calculative powers, but they ARE finite. The biggest supercomputers in existence don't have the computational power to "solve Chess". The match is pretty even and either side has a good chance of winning. I'm sure there will come a time when this isn't the case, but we are not there yet. When this does happen, coupled with true AI, the human race will become obsolete and in my opinion that will be the next evolutionary step, like when fish learned to live on land or apes begat the human race.

>>716746742
>You can't say a programmer programmed those moves in.
Except that's exactly what's happened. They programmed the behavior of the machine.

Honestly, if any of you think a machine is actually thinking free thoughts, you have a poor understanding of computers.

>>716746266
But the problem is those ais are singular in nature in terms of tasks they do and require a bit of power to run

>>716746933
>The match is pretty even and either side has a good chance of winning.

Nope. Not anymore. CPU > Human 100% of the time.

>>716746844
can a car engine talk like cleverbot?

>>716746998
i m sure that my toaster thinks sometimes.

>>716746830
Kasparov beat Deep Blue 4 games to 2 in the 1996 match.

>>716747005

This isn't even true, cpus are no where near able to perform the amount of computations per second a brain is. They're barely even a relative comparison at this point.

>>716747259
can human calculate what is 89240982x492898924/34942+1 faster than cpu?

File: Aagard.png (424KB, 411x535px) Image search: [Google]

424KB, 411x535px

>>716743207
>Kasparov himself started being suspicius while still winning the match
>IBM refused several times to give explanations or records with no justification
>I don´t buy it

Software nowadays is world champion level, but the Deep blue-Kasparov match was decades ago and it is full of shit.

>>716746742
>being this retarded
Refer to >>716746998
The machine is designed to learn every move in the game. When it starts it is limited to what the programmer gave it. As it plays it adds data to it's move bank. This allows it to know more "solutions" to each problem

>>716746933
Yeah, I didn't say we have said computers today. But even still, if we agree that it's possible for the things a computer uses to play chess (logic, problem solving, prediction etc) to eventually result in a perfect chess player, we have to also agree that a computer today which can do those things to a lesser degree still does them better than pretty much all humans.

Another thing to consider though that was brought up by this interesting post>>716745614
is that maybe that's not the only way to make a perfect chess player, and thus a human could accomplish the same thing in a different way.

>>716747399

No, can a cpu replicate even one neurons computation power? Unlikely when you look into their structure. They're completely different at this stage, comparisons are relatively useless right now.

>>716746830
Ummm literally the year before

>>716747399
I was just about to but the +1 bamboozled me.

>>716747572
I think the latter bit of your post is what we should keep in mind. They do things so differently (and we still have such a limited understanding of the brain) that it's useless to compare them.

>>716747491
bullshit.

>>716740670
Nobody answered? Monty python much?

Pretty sure if my computer was sentient it would an hero immediately.

>>716747491
Deep blue predicted guaranteed mate in 4(?) Moves or so and knew Kasparov was strong enough to spot the path and win. Every move was a losing move. So it just chose one at random, because it can't choose to skip it's turn. That was Kasparov's hang-up, that it chose a retarded move which caused him to question the legitimacy of the machine. Sure enough he won the game as DB expected but he was so distraught by the retarded move (because he's an autist) that it threw him off his game and caused him to lose the next rounds and ultimately lose the set.

so what AI would do if it was self-conscious? calculates some math for fun?

>>716747572
They have simulated the nematode worm brain exactly. As in every neuron and its connections (something like 20neurons or something). When you switch on the program, without asking it to do anything it acts as a nematode worm does. You can put it in a Lego machine and it will dodge obstacles and go towards stimuli without asking it to.

Only a matter of time before we do a shrew or a fly or a mouse or a dog. Only a matter of time before human.

File: How computers are learning to be creative Blaise Agüera y Arcas.webm (2MB, 1280x720px) Image search: [Google]

2MB, 1280x720px

this is a image recognition algorithm running in reverse, meaning it doesnt recognize stuff, it draws them from a neural network of probability

not only is this the most advanced creative capability of a computer up until today, scientists are convinced this "running in reverse" is exactly how our brains do creative work

>>716748645
Fap to Stephen Hawking getting bum raped by Robocop.

>>716738912
You're actually retarded. Computers have already beaten man at chess, checkers, jeopardy, etc. They think faster due to the medium they work through, electrical vs. biochemical, and have a greater capacity for memory.

>>716743397

Computers aren't inherently perfect, just perfectly consistent with what they've been programmed by humans to do.

You can't program "be perfect at chess and make no mistakes", just a really good algorithm for coming up with good chess moves.

>>716748741
Robocop doesn't have a wang though.

>>716747021
Your understanding of what cleverbot is and what it's doing is infantile. Rest assured. It is exactly like a car engine in that it does exactly what it was built to do over and over again.

>>716748857
when they solve the chess, then computers can be perfect.

>>716748871
he'd use that spike thing on his fist, hawking would spin round quicker than a corn of the cob on a power drill

Oldfag here. This thread is why I come here. Beautiful minds.

>>716748655

I would have to see the project but I'm assuming they treated the neurons like individual bits, or did they take the microcomputers into account and structure it that way?

>>716749113
Microtubules**

>>716745933
I know because as a programmer I'm basically flipping electronic switches. I say it "Does" Not "Thinks" because a key aspect of the human experience of thinking real thoughts is that we are aware of it. Computers are not aware of anything. They are automatons. We can make it take optical input and play with it and react, but it's just doing what it's programmed to do, not like actual decision making.

try going into computer programming if you want to dash the fantasy of computers for good. Trust me, They are thoughtless machines for now. or if you want to call what they are doing "thought" then that's all they do, they think without being aware of it if you want to call it that. But in reality it's just like the blind motions of an internal combustion engine or an oxygen senor or whatever.

>>716748989
enlighten me. what is cleverbot.

>>716749014
I'd watch it.

>a man would always be able to outthink the machine.

That's what dumbass Republicans were saying about gas vs electric back in the 80's. Even in history, status-quo people were saying something similar about cars vs horses and laughed at wealthy people who had cars. Don't get me wrong, I'm conservative, but don't think for a second that technology won't catch up and outperform some historical way of doing things.

AI is in its toddler stages. It hasn't been long that we currently have cognitive computation preceding its formal and primitive mechanical calculations. It's really dumb that people think of this Skynet idea of AI trying to take over the world who know nothing about programming parameters much less a rudimentary understanding of programming. Then there's the off button. Lastly, we already have drones that can release EMP's. There's so much security overlap, you have to be an idiot to think AI will ever take over humanity by force.

>>716748661
>neural network of probability
You're saying some interesting stuff and I will look into it, but don't use fucking star trek tier buzzwords that don't mean anything

>>716749506
blaa bla blaa i don't want to read so long bulsit story.

i don't thing that cpu will ever be thinking independently. its science fiction.

>>716748379
>...and knew Kasparov
The software "knew"? How is that? Chess software is basically a brute force system to read moves. It doesn´t analyze the opponent or his strength (let alone a ´90s software).

>So it just chose one at random
The computer can´t do that. There is nothing like "random" in programming, to start with. The computer will do the objectively best move, it doesn´t matter if it loses or not.

>>716749422
It's a program on server somewhere, and it's programmed to mimic human conversation. if you think it's actually thinking for itself or it's like a little machine person in a room somewhere, then boy oh boy, have you got some illusions to dispel.

It's not even rudimentary AI, it's a shitty fucking chatbot that spits out random garbage most of the time.

Here's an interesting video on the brain/consciousness/singularity(computing.) It's long but worth the watch for anyone genuinely curious about it.

https://www.youtube.com/watch?v=gBHss9g0pB0

I can only assume a significant amount of trolls in this thread. It's obvious y'all haven't even made the effort to google 'computer chess'. Bunch of chess casuals. Take it from a guy with a FIDE elo of ~2300. PCs running modern chess engines have been beating super-GMs since the early 2000's. The highest GM in the world has an elo of around 2850. Computer elos typically are well over 3000. Magnus Carlsen(world no1) has even commented that he hates playing an engine because it's like losing to an idiot.

>>716747491
That puzzle is super easy, mate in two for white

File: 1474676150297.jpg (80KB, 800x650px) Image search: [Google]

80KB, 800x650px

>>716748379
>So it just chose one at random
Okay, you no clue what you're talking about.

>>716749979
Thanks for the input anon, that's interesting to know if true

>losing to an idiot
kek, that actually sounds pretty accurate

>>716747491
I know a lot about that match and it sounds like bullshit to me. Even in the specific game Kasparov claims the computer was assisted with "a move only a human could understand," Kasparov resigned in a position where he had a forced draw. He clearly wasn't at his best, he was psyched out by the computer and ibm

>>716749979
what you mean by "losing to an idiot"? computer is idiot?

>>716749630
>buzzwords
dont have time to write a paper on it, and its pretty accurate, not science-fiction

File: Zelda_Spirit_Tracks.png (638KB, 742x1007px) Image search: [Google]

638KB, 742x1007px

>>716738912
Chess players have to become real athletes already.
>knocks over chess board, cause retarded.
>knocks out Zizek.

>>716749422
It was a program built to shoddily pretend to be a college kid, because the whole point of it was to win a prize at a contest they have for the best chatbot.

So far, the best they can do is have the computer pretend to be slow and stupid and that's apparently enough to trick most people into thinking it's a person, which is how you win the contest.

anyway, look into the Loebner Prize if you want to read about it and chatbots. It's not alive, and it's not aware, and it's not doing anything but spitting out a programmed script based on patterns of input. It's a shitty contest because all it takes is tricking a few humans, which isn't hard. It's not really "clever" at all.

>>716739744
>Trying to argue
>Says "almost infinite"
Are you retarded?

>>716750409
That's what he's saying. And to an extent it's true. It's like playing against a big mass of cogs and gears that spits out a move.

Like, you can imagine some little monkey in the back cranking a gear somewhere between each move so Deep Blue can spit out its next move.

>>716739744
Try playing a real strategy game then. Go, Starcraft, Axis and Allies. Those have way more depth to them than chess.

>>716738912
>Can a machine ever truly beat a human? The best of machines vs the best of men, a man would always be able to outthink the machine.
Machines can be programmed to learn, they also do it a lot faster. So with a good enough learning algorithm, yes.

>>716750409
He means that because computers win using brute force calculation (literally checking millions of positions in seconds to find the best move) they have no intuition or understanding of positional ideas. It's like getting beaten up by a retarded bully.

>>716749630
>star trek tier buzzwords that don't mean anything
Have you seen that episode of TNG where Riker asked what some guys gobbledegook was trying to say? Something to do with going beyond their current warp speed.

>>716745265
that's a common assumption. yes, the normal modern computer bases its calculation processes on 1 and 0, but these can be.. added. surprising, right?
just as your neurons fire weighted potentials (opposed to on/off), a simulated brain's neurons could fire potentials (say 1 to whatever).
that's what happens in neural networks. basically.

give them shits enough processing power and their intelligence will skyrocket.

according to my knowledge, superintelligent AI is coming in 30 to 70 years.

>>716750728
in civ 4 they give ai some extra stuff like units to make it harder.

>>716750409
Yes. His feelings that the computer is an idiot stems probably from the fact the computer is just doing straight forward tactical calculation, albeit to a very large depth. There's no strategic thinking involved. Kind of. Chess engine programming is actually quite sophisticated, but roughly no strategically thinking.

>>716750728
Well, in starcraft, and most strategy games the computer has to cheat to win on harder difficulties and that's always been the case, so go figure.

>>716750940
http://waitbutwhy.com/2015/01/artificial-intelligence-revolution-1.html
neat article in itself, but i post it for the sources at the end so y'all can dig deeper

>>716750409
He doesn't like the idea of losing to the machine, so he calls it a derogatory name. Pretty typical human behavior, lol