Massive large projects like this almost always end in utter failure.
I can think of quite a few successful ones between the Manhattan Project and the LHC
Even the IBM cat brain project failed to accomplish much.
This is a continuation of IBM's "cat brain" (Blue Brain Project), it's got a new name to reflect the fact it's no longer just IBM paying the bills. The reason it has been given taxpayer bucks is because the "cat brain" was very successful from a scientific POV. The main goal of the project has always been medical research, AI is a sub-goal.
Intelligence is much more complicated than a mere randomly connected neural network.
IBM's Watson convincingly disproves your hypothesis. Besides this project is based on anatomical correctness, it's a detailed physical model of a real brain for medical research via simulation, nothing random about it. It is hoped that creating such a model will give us new insights in how the brain works in the same way "numerical wind tunnels" have given us new insights into engineering.
My dad retired in the 80's, he was chief (mechanical) engineer at a large firm, computer simulation was just starting to appear in the industry. Today there is not a hope in hell of winning a major engineering contract without it. Computer modeling has revolutionized both industry and science since I left HS in the mid-seventies, there are no signs that revolution is losing momentum, in fact just the opposite.
I haven't read the singularity book, however it seems to me, our species may be heading down the same evolutionary path as ants, in that an ants nest can be considered a single intelligent organism (the Borg, if you prefer). Many people claim that the difference is that ants had no choice, but I doubt humans are totally immune to evolution just because our technology now allows us to air-condition our buildings like ants have been doing for millions of years.