Sunday, July 13, 2008

Computing's meaning of life - AI


Video courtesy of the Internet archive and thanks to Bruce Damer. For more go to the Internet archive

Yesterday I was treated to one of the most enlightening lectures I have heard in a long while from Bruce Damer talking at the Computer Society, Covent Garden, whom posited a powerful argument, which is also at the heart of his Phd.

His multifarious sources include the work of Karl Sims a computer whizkid at the time in 1994 at MIT who conducted a unique experiment which would simulate Darwin's theory of evolution.

Sims created virtual blocks and a soup of code which the blocks used to 'evolve'. Their task was to first swim through virtual water and then walk on land.

At each moment of their cycle they would run through a matrix of code and create a new virtual being, which would adapt to its surroundings, eventually learning how to paddle successfully through water and then adapt to mobility on land.

If that was 1994, what might be capable now in 2008 is in part Bruce's line of thinking.

Could we from nothing see how a virtual world evolves?

Bruce's quest
Bruce's work in what he's termed the evo grid is of such computing magnificence as well as so deeply philosophically penetrating that you need to be at one of his lectures to capture its essence.

I did however VJ his seminar and will be dropping him some mail both to say thank you and how I might be able to develop a film of sorts.

His previous works have included developing AIs/landing capabilites for NASA, physically collecting and documenting for his digibarn collection every computer ever made, and genesis Avatar development from the 90s.

There are some brilliant developments from virtual reality and AI, that almost appear astonishingly unreal, such as printing live tissues, using a 3d printer from a virtual scan, according to Dr Rachel Armstrong.

As a footnote, a couple of us had our own eureka moments thinking our own PhDs could raise their game, and that's one of the fascinating things about the Smart Lab.

It's eclecticism is also one of its strengths, giving people like me access to hear someone like Bruce both in lectures and within the smart lab environment, that I can't think how else I might ever have come by that opportunity.

No comments: