DARPA killer AI robots to 'participate in own construction'
Foolish meatsack! OF COURSE I need nuclear weapons!
You've got your robots which can make copies of themselves, of course. That's pretty scary - a runaway exponentially-multiplying machine horde, potentially able to overrun the human race in an eyeblink. But how much more scary would it be if you had a machine which could not only make copies of itself once complete, but could also participate in its own construction while it was still being built?
Really quite a lot more scary, we say here on the Reg mechanoid-armageddon desk. That's why we would never be suitable for employment at DARPA, the famed US military tech bureau where they care not what deadly man-eating monster worm plagues may be unleashed during tests of a new laser-powered can opener.
Yes, it's true. DARPA have now expanded somewhat on their intention to initiate a programme called Self-Explanation Learning Framework (SELF), which they explain thus:
DARPA seeks to construct systems that can participate in their own construction... The system might know the requirements for various tasks in its repertoire, and it may try to perform those tasks to verify functionality.
We particularly liked that last bit. One should bear in mind that Dr Mike Cox of DARPA has already said that SELF could be placed in charge "in the near term" of heavily armoured, hideously beweaponed main battle tanks or strike planes laden with blockbuster bombs.
"Tasks in its repertoire", then, might include "destroy all moving objects within 100 miles not designated as 'friendly'" or "mount an immediate armoured assault on Beijing, regardless of nuclear response". The prospect of the software unilaterally "trying to perform those tasks to verify functionality" doesn't seem reassuring.
In any case, it seems plain that building a system focused on "high-level cognition" which can "participate in its own construction" will be fraught with difficulty. You might have a notion of what you'd like it to be, but it will have its own ideas. By definition, one would have no firm picture of just what would be unleashed upon an unsuspecting world at the end of the construction process - but you would know that it'd be potentially able to make more of itself, or indeed repair itself if it got damaged.
It might also be rather difficult to stop the process of building the SELF, once it had advanced beyond a certain point. Frankly the only way to be sure it can be stopped would seem to be to stop it now.
But that's not the plan. Dr Cox and his DARPA chums are holding an industry day for those interested in
building starting the building of SELF. The details are here in a pdf, and a website is given where attendees can make known their "special needs" (for instance supervised day release from the Asylum for Troubled Scientists). ®