Grand Challenge

Somehow I lost track of time, but the Grand Challenge results are fresh news for me.

I have some personal interest in the Red Team not only cause my company helped pay for it and donated equipment, but also cause it was driven by those nuts at CMU.

I don't really agree with the terms of the competition though. It is still trying to fulfill that old stupid dream of GOFAI (good old fashioned AI). I mean, I don't care if my car drives itself 100% in 10 years. I would rather it be able to drive itself 95% of the time tomorrow, which is achievable today. But this is assuming that we aren't afraid to embrace this human-computer symbiosis, and instead continue to divide humanity in grotesque ways between personal and impersonal interfaces (externalized, embodied intelligence versus a social, distributed cognitive intelligence).
It also assumes that the holy grail is not achievable in 10 years... so for 35 years it has not been... and so why not introduce HUMAN intelligence (eh, in situ you goons) into the mix? The everyday computer user's intelligence far exceeds performance of the best GOFAI (in most domains). We should be bridging these intelligences as well as decomposing and engineering them from scratch. Research dominates the one end of the spectrum however... I'll just get back to work now and stop being contrary.

Your PC sucks! More about Red Team