Google's Deepmind AlphaZero chess AI is apparently learning itself human intuition and creativity

[Master Disaster]

3 minutes ago, Shally said:

I don't get why it's doing this, maybe I read wrong but it's sooooooooo much further ahead in chess than everything else so why would it bother to learn human creativity? Is it a side project for the dev team? 

And that's exactly what scares me about this entire project. Afaik the team hasn't had any input into this things learning decisions, they gave it logic, set it loose and it seems to have chosen to learn "intuition and creativity" itself.

 

I would be interested to know how they handled its winning logic, does it just want to win or does it need to win?

 

If it's the latter then the thing showing signs of self improvement is a scary concept.

[LAwLz]

8 minutes ago, Master Disaster said:

I would be interested to know how they handled its winning logic, does it just want to win or does it need to win?

 

If it's the latter then the thing showing signs of self improvement is a scary concept.

What do you mean by "does it want to win or does it need to win"?

It's programmed so that it tries to win. It has no concept of "wanting" vs "needing". 

[Master Disaster]

1 minute ago, LAwLz said:

What do you mean by "does it want to win or does it need to win"?

It's programmed so that it tries to win. It has no concept of "wanting" vs "needing". 

I'm more talking about it's psuedo need for information to self improve.

 

Have they designed it so it's happy to draw and maybe not gather information or have they designed it to always want to get something out of any given situation?

 

I realise im using human constructs to describe a machine but I don't know any other way to say it.

[Doobeedoo]

Impressive. I like the take on StarCraft can't wait to see more on RTS side of things how it improves more. 

[BuckGup]

14 hours ago, Trik'Stari said:

Has any of these people thought that maybe we should slow down a bit?

 

Automation stands very likely to put a third of the workforce of the developed world, out of a job. Less than 1/5th-1/15 of those (in my mind) is likely to find a new job repairing the automation machines. And how long before those people are replaced by machines?

 

I wouldn't mind, if we were seeing huge breakthroughs in "putting my brain in a jar on top of a robot so I can live forever".

If you talk to people developing AI you will realize these news stories sound scary but AI is no where close to becoming self aware or take over the world. Also interpreting AI data is damn near impossible so a lot of the time a computer will do things that we have no clue why and some news outlet will blow it up saying AI MADE A LANGUAGE ENGINEERS COULD NOT READ WAS PLOTTING SOMETHING

[LAwLz]

2 hours ago, Master Disaster said:

I'm more talking about it's psuedo need for information to self improve.

 

Have they designed it so it's happy to draw and maybe not gather information or have they designed it to always want to get something out of any given situation?

 

I realise im using human constructs to describe a machine but I don't know any other way to say it.

I think you're trying to, as you said, use human constructs and emotions and applying them to a machine. They don't think that way. They follow instructions and have no concept of "need" or "want". It's kind of like asking if hammers have a want or need to hammer in nails. 

[ZacoAttaco]

15 hours ago, Master Disaster said:

I hope they know how to handle these machines they're creating.

That’s the issue right there... they’ve shown a lack of control with the YouTube algorithm, it’s something they let loose and have to clean up the mess after it. 

 

So no, I don’t really think they know how to control what they’re creating, because it continually innovates on itself, it’s constantly changing and evolving. 

[FPSwithaWacomTablet]

5 hours ago, LAwLz said:

Spotted the Luddite.

 

People said the same thing when "spinning Jenny" was first developed.

"If we no longer need people spinning threads 10 hours a day those people will be out of jobs, therefore this machine is bad!".

 

Machines taking over jobs has historically meant that people can spend more time on creative or more skill intense jobs. I think the problem is that people view available work as a finite amount and as machines start doing some tasks, we won't have enough work to keep people busy.

If we look at history, we can see a clear trend that as machines have taken over jobs, peoples' time have been freed up so that they can spend more time on doing creative or otherwise higher skilled tasks.

Someone who worked 10 hours a day spinning threads can now spend 8 hours a day coming up with designs for clothes to make with the thread for example.

 

The cost of production has also dramatically dropped as industrialization has taken place, which has lead to higher salaries and shorter work days (which again, means more time for fun and creative stuff).

I will also quote something my grandmother used to say: "Idle hands are the devil's hands".

 

While people will be freed to pursue other things in life, it's also possible that people can become revolutionaries. Not everyone would have the predisposition to pursue other careers, and therefore be even more displaced from the job market as the occupations dry up with time. That is usually a driving force in revolutionary movements, which is not an ideal position a country would like to be in. 

[Syntaxvgm]

16 hours ago, Trik'Stari said:

Has any of these people thought that maybe we should slow down a bit?

 

Automation stands very likely to put a third of the workforce of the developed world, out of a job. Less than 1/5th-1/15 of those (in my mind) is likely to find a new job repairing the automation machines. And how long before those people are replaced by machines?

 

I wouldn't mind, if we were seeing huge breakthroughs in "putting my brain in a jar on top of a robot so I can live forever".

sure we can slow down, and be left behind by everyone else doing it. 

[vorticalbox]

8 hours ago, RejZoR said:

If you can predict all movements through calculations in a fraction of a second on every turn, you don't need intuition

That's not how this ai works, it does not calculate all the moves it learns to play better by making mistakes just like you or I would.

[Trik'Stari]

16 hours ago, LAwLz said:

Spotted the Luddite.

 

People said the same thing when "spinning Jenny" was first developed.

"If we no longer need people spinning threads 10 hours a day those people will be out of jobs, therefore this machine is bad!".

 

Machines taking over jobs has historically meant that people can spend more time on creative or more skill intense jobs. I think the problem is that people view available work as a finite amount and as machines start doing some tasks, we won't have enough work to keep people busy.

If we look at history, we can see a clear trend that as machines have taken over jobs, peoples' time have been freed up so that they can spend more time on doing creative or otherwise higher skilled tasks.

Someone who worked 10 hours a day spinning threads can now spend 8 hours a day coming up with designs for clothes to make with the thread for example.

 

The cost of production has also dramatically dropped as industrialization has taken place, which has lead to higher salaries and shorter work days (which again, means more time for fun and creative stuff).

You're making the assumption that 1 job lost = 1 easier job created.

 

Which is simply, historically, not true by any measure. Historically speaking, automation has lead to poverty, increasing prices despite decreasing costs, etc. Go visit Detroit some time and see how run down it is. Automation killed areas like that. Just like cheap labor overseas killed areas like the city I used to live in (Danville, Virginia). The big plants close down, smaller business open up sure, but it's still an overall net loss.

 

I never said stop, I said slow down.

[Bigbootyjudy]

Hmmmmm, yes, now these robots will take over the world, BY ESCAPING THE GAME OF CHESS ADN SENDING A ROOK TO THE U.N.!!!!!

[yian88]

Good it can then kill us in creative ways.

[LAwLz]

7 hours ago, Trik'Stari said:

You're making the assumption that 1 job lost = 1 easier job created.

More or less, yes.

 

7 hours ago, Trik'Stari said:

Which is simply, historically, not true by any measure. Historically speaking, automation has lead to poverty, increasing prices despite decreasing costs, etc. Go visit Detroit some time and see how run down it is. Automation killed areas like that. Just like cheap labor overseas killed areas like the city I used to live in (Danville, Virginia). The big plants close down, smaller business open up sure, but it's still an overall net loss.

The problem there is that the jobs moved to different places.

The automation of car production may have killed a lot of jobs in the Detroit area, but how many jobs do you think were created by more and cheaper cars? I don't have any exact numbers, but if we look at it on a global scale I am fairly sure it's more than those who lost their jobs in Detroit.

 

Detroit have around 670,000 citizens.

3.5 million people in the US alone are employed as truck drivers.

Uber have over 3 million drivers (plus Lyft, and Didi).

On top of all the people who now have access to work that they need cars to get to.

 

Even if every single person in Detroit used to work for the car industry and are now unemployed, the number of work opportunities is still higher.

 

 

And since we're talking about cars, my guess is that you will bring up driverless cars.

A study by KPMG for the SMMT estimates that by 2030, driverless cars in the UK will have created 320,000 new jobs in the UK.

 

 

7 hours ago, Trik'Stari said:

I never said stop, I said slow down. 

If you believe that it will result in loss of jobs, then what difference does it make if it happens now or in 10 years? The result will be the same.

[Livinloud]

From what was said up above AlphaZero Chess as only beat SF8(stockfish) and I believe 9. No test as of yet of SF10. AlphaZero first started with GO which has already beaten the 1# player. Chess I would say still as a way to go before being number 1. Currently there is a project going on LeeLa chess is using AlphaZero AI that you can currently help them with using your CPU or GPU got calculating moves or even just play against. But you will get like 14knodes on a 1080ti were as SF is in the millions.