Well put. The Open Philanthropy folks take a broadly similar view of the probabilities, both of when strong AI is likely to arrive and of how much positive influence per dollar they can have on how things turn out:

http://www.openphilanthropy.org/blog/potential-risks-advance...