The role of central planning in China will certainly allow to get a lot of things done. It is true about bridges here in America, a lot of infrastructure is in a bad shape or just absent, which is easy to see traveling the country.The biggest issue with AI in the USA will be building the infrastructure. We can't even rebuild a bridge after a few years, that was built over 100 years ago. How will we ever build the 40 to 70 estimated new nuclear power plants needed to power the massive data centers, needed for the AI advances?
China can build this, partially because they have a harder working/higher IQ population v. the USA and mostly because the CCP and their fascist style govt. will make it a top priority and see that it is done.
AI infrastructure is financed mostly as private investment projects in the US, the key difference with bridges, so they will find the money for these projects, which can be created out of thin air by loans, as soon as there are loan takers interested in stealing the loot
One problem is how expensive everything is here. Labor, parts are expensive, people having to pay 50k for health insurance and thousands in mortgage monthly (pay up to racketeers aka medical, insurance and construction cartels) while paying off their student loans, can only produce very expensive things. It is a spiral of infrastructure cost increases, as at every stage someone needs to bite a big chunk of profit off and this pyramid builds up on overpriced base. Such expensive AI needs to be extremely profitable to be paying off, but who will be paying into those big profits, just to cover the costs even - already financially strained "consumers", who are losing their jobs to AI as it expands? The math does not addup. I just hope this won't end up in taxpayer funded bailout 2008 style because of "strategic importance" of overpriced AI stuff.