1 billion dollars earmarked for whole brain simulation makes it seem a lot more likely that we'll brute force a naive version of AI well before we have the ability to prove any kind of friendliness. If that AI is seeded by the simulated brain of an actual human though... who knows. I'd like to think that if it were my brain and at some point I became singularity-scale intelligent that I wouldn't create a horrible future for humanity (by our present day perspective) but it's pretty hard to claim that with any confidence.

They are very careful to not mention certain things in their video: nothing involving the computational nature of the human identity / "consciousness". Nothing we would call "whole brain emulation".

I have little doubt that what they choose to mention, and what they choose not to mention, in their video is driven more by signalling considerations than by what they consider to be technologically (in)feasible. Can they do what they claim? Create a model of the human brain detailed enough to obtain non-trivial information about brain diseases and computational processes used in the brain, but coarse enough to not have moral implications? Of course I do not know the answer to this question.

Kurzweil, who is widely agreed to be rather optimistic in his predictions, predicts

By the mid 2020s, it is conservative to conclude that we will have effective models of the whole brain

See this comment