Re: I agree with Musk
"Is there something inherent to biology (carbon) that makes it superior to artefacts (silicon) in nurturing intelligence".
Biology does not need software. Biology can adapt itself without invoking a huge bureaucracy, scores of developers, diversity committees and agile facilitators.
I also think it is very likely that biology is not the only way to build / grow an AI and maybe not the best way either. However, the writing of millions of lines of brittle code and running it on foot-ball sized areas of silicon is at best only a modelling effort, nothing "radically new" will likely come from all that effort because there is some physics missing that biology has and silicon is missing.
I believe that for an AI to really work as in being genuinely intelligent and not just a good simulation of intelligence, the computing has to happen inside of the fabric, not in the code. In a way similar to biology where genes describe the general layout of the "intelligence factory" and once that is built, the intelligence happens on it's own without any code being needed.
There is an interesting approach called "Reservoir Computing" where one does use materials directly as a computing fabric. It seems like "computing" is another "physics"-property of normal matter. This being the case, "all" one needs to do is to arrange the matter appropriately so that the physics can work. This is what biology figured out and the silicon+code approach per-design cannot do.
Now, "They" probably want exactly a clever simulation of intelligence rather than true intelligence because one can control and limit the simulation. "Their" vision is like seen in those SciFi movies where they have robots and AI while also having poor people and serious class divides. The real thing will be much harder to control and exploit. It might kick off and become an immortal Martin Luther King or Digital Jesus curing us from disease and poverty and then where would "we" be when rich or poor is just a matter of personal preference?