>>11134139He is overly optimistic in his assumptions.
I.e.: At one point he argues that he does know that the current sillicon based processors are clearly approaching the top of the performance "s-curve", and thus, the age of exponential performance and computational power is ending ("Moore's law in a nutshell").
But he also states that history of human technology is not just one s-curve for each technology. That different technologies arise to overlap and step-in, just as older technology reaches it's limits.
In essence, he's saying it does not matter we are reaching the limits of what our current computers can handle, because humans will surely come up with new technology (cough, cought quantum) that will pick up where silicone left off.
That is false.
The only technology we know of, that could potentially replace silicone, is carbon nano-tube based processors. And those are nowhere near ready for application in real world. And quantum computing is not a replacement for existing computing. It merely augments it with new types of computation.
tl;dr: He's projecting future based on past performance, without actually explaining how we're gonna get there.
