Linked by Eugenia Loli on Mon 28th May 2012 03:53 UTC
General Development FuriousFanBoys interviews Ben Goertzel regarding Artificial Intelligence. Ben started the OpenCog project (an open sourced AI non-profit), acts as an adviser to the Singularity University, and currently bounces back between Hong Kong and Maryland building in-game AI.
Thread beginning with comment 519659
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[3]: Memetics
by zima on Mon 28th May 2012 06:09 UTC in reply to "RE[2]: Memetics"
zima
Member since:
2005-07-06

But I would argue humans don't think like a human yet either. I've never met any people in significant numbers that uses the whole of human experience in their "intelligence". They mostly use a very rigid subset [...]
It brings up another question: why on earth would we consider an AI to be insufficient if it doesn't match a human? They exceed humans in many other tasks already.

I think people go even further - they tend to expect from an AI to beat exceptional human, maybe even "the best" one...

...while AI is really more about being better than average human, inexpensively mass-producing and distributing its expertise. That is sufficient to bring improvement to the world.

Sure, AI defeated chess world champion only in 1997 - but I suspect it could beat most humans quite a bit before that.
(heck, I remember that for me, then a small kid, some C64 chess program was a challenge ;) )

Edited 2012-05-28 06:09 UTC

Reply Parent Score: 2

RE[4]: Memetics
by kwan_e on Mon 28th May 2012 06:14 in reply to "RE[3]: Memetics"
kwan_e Member since:
2007-02-18

...while AI is really more about being better than average human, inexpensively mass-producing and distributing its expertise.


This is kind of why I think artificial intelligence has already been achieved. It was already achieved the first time ELIZA successfully trolled the participants. The rest is just people making excuses to hide away from the fact that most people are stupid (in the best sense of the word) and people just vary in their abilities and their proficiency in those abilities.

Intelligence is not the holy grail here. It is consciousness that we're really chasing. We've had the benefit of having physical bodies, and I don't think artificial consciousness can really proceed unless it has physical bodies with which to evolve along with.

Reply Parent Score: 2

RE[5]: Memetics
by zima on Mon 28th May 2012 06:28 in reply to "RE[4]: Memetics"
zima Member since:
2005-07-06

With chatbots, sometimes I think that a much fairer Turing test would involve testing humans who must communicate in their non-native language (with a representative spectrum of proficiencies)

After all, not only that would approximate what the AI must do, it's also much more representative of random human-human communication... (you don't know the language of, can hardly communicate with strong majority of humans)


And BTW physical bodies, are you sure you have one & have you heard about simulation argument? ;)

Reply Parent Score: 2

RE[5]: Memetics
by cfgr on Mon 28th May 2012 13:26 in reply to "RE[4]: Memetics"
cfgr Member since:
2009-07-18

AI has gone much further than that, and 20 years ago.

Computers are perfectly capable of learning by trial and error to beat humanity hands down. Gerald Tesauro developed a machine that can teach itself to play backgammon.

It does so by playing against itself and exploring policies, not moves, and by punishing/rewarding those policies for how well it is doing. It's called reinforcement learning. The code did not contain any strategies, only the game rules. After thousands of games against itself, it managed to beat average players, Tesauro included.

Then they let it learn from analysed situations from a database (i.e. like from a book) - again nothing coded. After this it could defeat the world's best players and in fact changed the way grandmasters play backgammon.

It's pretty awesome.

http://www.research.ibm.com/massive/tdl.html

Reply Parent Score: 2

RE[4]: Memetics
by MOS6510 on Mon 28th May 2012 08:17 in reply to "RE[3]: Memetics"
MOS6510 Member since:
2011-05-12

(heck, I remember that for me, then a small kid, some C64 chess program was a challenge ;) )


Not the C64 chess program I had, it didn't check for illegal moves from the user and you could add pieces whenever you liked.

There also was a chess program for the ZX81 that ran in 1 kB of memory.

http://users.ox.ac.uk/~uzdm0006/scans/1kchess/

Reply Parent Score: 2

RE[5]: Memetics
by zima on Mon 4th Jun 2012 23:28 in reply to "RE[4]: Memetics"
zima Member since:
2005-07-06

So, it made me wonder which I had... found a list in the most straightforward place http://en.wikipedia.org/wiki/List_of_chess_software (meh, too much clicking / often lacking pictures), then a really nice historical outline: http://www.andreadrian.de/schach/index.html (pictures!)...

...some other programs there seem at least as ~impressive as that 1k ZX one (which is also included of course), for example http://en.wikipedia.org/wiki/Microchess for KIM-1

Or http://en.wikipedia.org/wiki/Video_Chess for Atari 2600 ...with some curious GFX acrobatics and 128 bytes of RAM (yeah, 4K ROM compensating it somewhat; still, think about it, even keeping the state of the board must take non-trivial part of RAM - from a link in Wiki art about ZX 1K: http://www.kuro5hin.org/comments/2001/8/10/12620/2164?pid=22#24 ) - I must toy around with development on that machine sometimes ;)

Mine was almost certainly some version of http://en.wikipedia.org/wiki/Colossus_Chess (unless there was some other lookalike...). And without such easy illegal moves, for sure - actually, perhaps with quite decent playing strength, judging from comments on lemon64 links.
Plus, http://en.wikipedia.org/wiki/Sargon_(chess)#Sequels "Even though chess programs of the time could not defeat a chess master, they were more than a match for most amateur players." is soothing ;)
(because when I wrote that Colossus was a challenge, I didn't mean that I couldn't beat it - and me being an amateur, never "formally" trained in chess; who knows, perhaps those Colossus matches from almost 2 decades ago taught me something... either way, other non-trained humans never seem to be a match for me - and, on the few occasions I played with chess-trained human players... they of course won, but it supposedly took them more time and effort than is typical, when confronted with other total amateurs)


And generally, thinking about it made me realize one curious thing - how relatively inexpensive home computers of the early 80s seem, vs. what came later (PCs of the late 90s, most notably)

Reply Parent Score: 2