The DARPA IPTO Department and A.I.

What do you all think about the most recent solicitation for Cognative Systems from DARPA IPTO?

Cognitive Information Processing Technology:
www.darpa.mil...

Some of the many point raised and are solicited to be answered with the winning research efforts:

"How might a cognitive system learn the salient things from each experience it has and later use what was learned in an appropriate way to interpret and successfully cope with new situations? How can it find the right remembered experiences to apply to each new situation? "

"How can we build systems that effectively keep an eye on themselves? How can an artificial reflective system operate in real time? Is there virtue in allowing reflective processes direct access to internal structures and processes, i.e., true introspection? "

"How do we build systems that can deal with proper prioritization of �standing orders,� given complex and even conflicting goals?" (The child is innocent, I am programmed not to kill innocent beings, however I was order to kill the child).

How can we create teams of agents whose �collective IQ� is at least as high as the �IQ� of the components? Why is group decision-making so poor and how can we improve on human performance in an artificial setting? (Can we say - the Borg?)

This is what Darpa is looking at.

I'd like to know your thoughts on ethics of A.I. in this case and the plausible future of A.I. within the US military as a result of this specific research solicitation.