Some Perspective

Deep Blue and Potato Peelers

Come to think of it, the win of Deep Blue was a milestone for the field of A.I. But definitely not for pure computer intelligence, as we've just discussed. Deep Blue's win was a major step towards assisting human intelligence with computers.

Deep Blue is a lean, mean chess machine, designed by experts for the single task of anihalating Gary Kasparov. Many chess experts are convinced this is not a random choice, as Kasparov is not the best anti-computer player, unlike other players with more varied styles of positional chess. So in essence, this tool was created for a specific purpouse, and optimised to shreds therefore. Deep Blue is just like a well designed potatoe peeler (the metaphore makes me chuckle everytime I think of it). With the assistance of a human hand, it can peel potatoes extremely well.... even better than any human alone has ever peeled potatoes. However, give it a carrot and the human hand will have to adapt to peel it.

Assisting Human Intelligence

This is reminiscent of the current state of A.I. in general. We seem to be, conciously or not, setting out to use computers to assist human intelligence. The combination of natural and artificial intelligence is, in the short term at least, more promising than either of the two alone.

A very recent article discusses this, starring once again Ray Kurzweil: The New Future.

More Information

There's a plethora of information out there on the web about this very topic. It's naturally captured the attention of a major part of any journalist at some stage, not to mention pretty much anyone involved in A.I.

The first place you might want to check out is IBM's Chess Page, where you'll find the logs of the games and many interesting facts about deep blue. Some of the guest essays are more interesting than others, but most provide very good reading.

After that, you'll be ready to tackle more philosophical work, things like:

• How Intelligent is Deep Blue? - Argues that deliberative reasoning does not need to be concious to be classed as intelligent, and that the lack of flexibility in Deep Blue makes it un-intelligent.
• The Meaning of Deep Blue's Victory - Goes into an analysis of Deep Blue's functioning, comparing it with human intelligence. It then turns into nasty stuff: analysing the dangers (oooh!) and unpredictability of A.I. in the near future... hmmmm.
• Rethinking Deep Blue: Why a Computer Can't Reproduce a Mind - Talks about how great Deep Blue is at chess, saying it's invincible! I'm yet to be convinced, see my thoughts in the previous pages... The interesting part is the discussion of the frame problem, mentioning how deliberative thought can have its draw-backs. The conciousness of A.I. section is starting to become outdated though, as much successful work on emotions and sensations has been done since 1997.

And finally, if you think Chess is a hard game, meet Go! Go is known as chinese checkers, and has a search-tree that is absolutely huge compared to that of chess. We'll no doubt look into this at a later date

Remember you can visit the Message Store to discuss this essay. Comments are always welcome!. There are already replies in the thread, why not join in?