A New Force In The Universe! AI has arrived.

Sort:
RMChess1954

When Gary Kasparov lost to Deep Blue he lost to a database of knowledge gathered by mankind for over 500 years, administered by man made algorithms that applied that knowledge. Today Google's DeepMind learned the game of chess and a few hours later had created it's own knowledge base. Then it crushed the current world champion chess engine. Now the computer DeepMind was running on may have been 1000 times faster then the one Stockfish was on but still the implication is a whole new way of computing. One in which the computer may find new approaches to the problems we face. It may be better, much better at things like how to create a fusion reactor, or how to cure cancer. Think about that for a minute.

blueemu

I have to wonder why Google didn't arrange a fair test, using equivalent hardware... after all, it's not an earth-shaking revelation to learn that a program running on a super-computer out-performs a program running on conventional hardware.

RMChess1954

Stockfish was running on hardware that would place it's ELO above 3000. A very strong opponent. Deepmind running on a supercomputer but without an opening database. Using algorithms it created itself. It's not like two current engines facing off. The supercomputer was needed to run the AI not calculate variations. 

blueemu
RMChess1954 wrote:

Stockfish was running on hardware that would place it's ELO above 3000. A very strong opponent. Deepmind running on a supercomputer but without an opening database. Using algorithms it created itself. It's not like two current engines facing off. The supercomputer was needed to run the AI not calculate variations. 

Yes, I'm aware of this. But better hardware would have improved Stockfish's performance, making it an even more formidable opponent... and correspondingly more impressive if AlphaZero won consistently.

IpswichMatt
RMChess1954 wrote:

It may be better, much better at things like how to create a fusion reactor, or how to cure cancer. Think about that for a minute.

Don't hold your breath. I suspect it may be good for driverless cars and other things that take jobs away from people and make money for google. If it helps find a cure for anything or helps with fusion I'll eat my fckuing shoes.

hairhorn

And seriously, what's up with the one minute time limit? 

szaszzo66
RMChess1954 írta:

Stockfish was running on hardware that would place it's ELO above 3000. A very strong opponent. Deepmind running on a supercomputer but without an opening database. Using algorithms it created itself. It's not like two current engines facing off. The supercomputer was needed to run the AI not calculate variations. 

Well, as I understand the reports, it calculated variations, too -- both while learning and also during the match.

RMChess1954
szaszzo66 wrote:
RMChess1954 írta:

Stockfish was running on hardware that would place it's ELO above 3000. A very strong opponent. Deepmind running on a supercomputer but without an opening database. Using algorithms it created itself. It's not like two current engines facing off. The supercomputer was needed to run the AI not calculate variations. 

Well, as I understand the reports, it calculated variations, too -- both while learning and also during the match.

Yes I wasn't very clear about that. Humans also calculate variations. Deepmind was not calculating to near the depth that Stockfish was. This says that it doesn't get it's strength from the depth of it's search but from the way it evaluates the position. Stockfish beats Komodo and the other engines by being faster, and looking deeper. Deepmind won by looking better with it's own knowledge. If the paper holds up to the review process. This is all depending on what they are saying is true.

me_roma

AI's empowered with learning algorithms. If that does not scare the s**t outta you, I don't know what will..

*200 years later* "You are in the matrix.."

RMChess1954
me_roma wrote:

AI's empowered with learning algorithms. If that does not scare the s**t outta you, I don't know what will..

*200 years later* "You are in the matrix.."

Donald Trump is our President. AI is not scary compared to that.

sammy_boi
chiefonion wrote:

I don't understand why people are so amazed by this. It had to play a bunch of games before it got stronger. Just like a human. If you gave AlphaZero a Facebook account and told it to pick up the kids from school, buy dinner, gave it access to Youtube videos, etc... It wouldn't be as strong.

 

The answer is quite clear, humans are not playing chess as strong because of lifestyle choices. AlphaZero had no distractions, no other thoughts creeping in. Humans could be just as powerful if they do the same.

It doesn't sleep, and only need 4 hours to be as good as stockfish.

So even if it had to pick up the kids and take a poop, humans wouldn't be as good.

Even though I shouldn't be saying this, because you're only trolling.

me_roma
chiefonion wrote:

I don't understand why people are so amazed by this. It had to play a bunch of games before it got stronger. Just like a human. If you gave AlphaZero a Facebook account and told it to pick up the kids from school, buy dinner, gave it access to Youtube videos, etc... It wouldn't be as strong.

 

The answer is quite clear, humans are not playing chess as strong because of lifestyle choices. AlphaZero had no distractions, no other thoughts creeping in. Humans could be just as powerful if they do the same.

That's a daring hypothesis. But it did remind me of a statement that a normal human being utilizes <1% of the true potential of his/her brain. Anyway, I do not think you can compute the number of prime numbers between 0 and 1000 within a couple of seconds, even if you are free from all distractions. But I can provide you a simple code and a compiler that can do it in a flash.

RMChess1954

 Example of a Troll. The poster chiefonion joins chess.com. Plays no chess, just jumps on the forums and starts posting stupid nonsense. But that is ok as long as he/she/it remains civil. I'm just pointing out what many of us know. That there are trolls on here who don't even know how to play the game and never will. They just join to spew cr*p all over the forums.  

szaszzo66
me_roma írta:

That's a daring hypothesis. But it did remind me of a statement that a normal human being utilizes <1% of the true potential of his/her brain. Anyway, I do not think you can compute the number of prime numbers between 0 and 1000 within a couple of seconds, even if you are free from all distractions. But I can provide you a simple code and a compiler that can do it in a flash.


Maybe a bit more than this: like 8%, some of us maybe even 12%. But the human brain operates at TeraFlops/sec speed! So, we are not slow, just distracted and/or multitasking -- with no transparency whatsoever! happy.png This is not transparent at all to our consciousness... we don't realise it and most of us even don't know it.

me_roma
szaszzo66 wrote:
me_roma írta:

That's a daring hypothesis. But it did remind me of a statement that a normal human being utilizes <1% of the true potential of his/her brain. Anyway, I do not think you can compute the number of prime numbers between 0 and 1000 within a couple of seconds, even if you are free from all distractions. But I can provide you a simple code and a compiler that can do it in a flash.


Maybe a bit more than this: like 8%, some of us maybe even 12%. But the human brain operates at TeraFlops/sec speed! So, we are not slow, just distracted and/or multitasking -- with no transparency whatsoever! This is not transparent at all to our consciousness... we don't realise it and most of us even don't know it.

Thanks for the information. This is interesting, I'll look into it.

szaszzo66

Just imagine the brain as an industrial process-controlling computer or CPU. It has to control gazillions of things in the body, e.g., blood pressure, heart rate, electrolythic (I might have misspelled the word) balance etc and also our sensory interface with the outside world (meaning the five senses, which mean many-many stimuli per second).

IpswichMatt

The idea that humans only use about 10% of their brains is a myth

szaszzo66
IpswichMatt írta:

The idea that humans only use about 10% of their brains is a myth

Care to elaborate on this?

IpswichMatt

If you google it you'll get many hits saying it's nonsense. One of the more interesting is the faculty.washington.edu site which talks about the origins of this myth

szaszzo66

Well, maybe this 10% pertains to its conscious activities...