The U.S. needs a shipbuilding revolution

At the very list ships built in Italy, (NATO partner), Japan, South Korea (close allies with a ship building industry) should be allowed. Probably we should allow countries like Kenya, Vietnam, Chile (random non-nato countries that don't have ship building but could and seem like places that we want to encourage to become closer to us).

I would argue that it isn't only possible, but on track to arrive sooner than most people realize:

* AI models are steadily continuing to improve in capabilities and efficiency

* Massive investments are being made in scaling up AI infrastructure (see Stargate and xAI Colossus)

* Tesla expects to produce a few thousand Optimus robots this year and use them for some level of internal production workload, meanwhile Hyundai has acquired Boston Dynamics with what I can only assume is a plan to take its tech out of the research labs and commercialize it at scale

* Aside from all the other recent and ongoing advances in energy tech and infrastructure, production fusion power is coming; if you take sama-backed Helion's word for it, they may be fulfilling a contract to deliver it to Microsoft as soon as 2028 (knock on wood)

Add all that together, and it's not difficult to see a trend that converges on a rapid massive expansion of global and particularly US manufacturing output kicking off within the next decade or two. As soon as the hardware and software are good enough for robots to outcompete average unskilled human laborers at most tasks on cost and quality, expect fully automated assembly lines to start pumping out humanoid robots 24/7, which will then be put to work 24/7 on any number of manufacturing and construction projects with logistics based around autonomous vehicles.

The overhead of US labor cost and safety regulations will become moot with machines doing the work, while our abundance of resources and first mover advantage on AI will give us a big headstart over the rest of the world. Meanwhile, our low population density means we'll have a ton of empty land to build on and a population size that will make UBI payments comparably easy. In that scenario, eclipsing 2025 China's shipbuilding capacity will be the least of our concerns. Whoever wins the AI race wins global hegemony, and right now that race is America's to lose.

All of which is to say, there's a reasonable argument that America is currently sitting at a firm local minimum in strength and prosperity, which conversely means that China is plausibly approaching a ceiling on its own relative military and economic power for the foreseeable future. If that is the case, it means that the next decade or so may be an exceptionally high-risk period for Taiwan. However, it also means that competent US leadership would throw everything it has at a defense of Taiwan in the event of an invasion; irrespective of any fabrication capacity that may end up built out in the US, allowing a Chinese takeover of the main TSMC facilities would be surrendering far too great a strategic asset in the AI race. That being the case, while Chinese leadership may or may not agree, I would argue that the rational move on China's part would actually be to give up on Taiwan and focus on investing heavily in SMIC and other fronts of the AI race. Invading would at best yield a pyrrhic victory, at worst yield an expensive defeat and burn a bridge with the people of Taiwan for generations. The right move would be to put aside the short-term economic gambit and nationalistic fervor, and instead lay out a roadmap for a possible future peaceful unification or alliance by proving themselves to be a good neighbor over time.

China and Russia are already expanding as we speak, being allied in the invasion of Ukraine

> > get good at things by doing them repeatedly and optimising

> In an environment where information is shared freely and widely.

The whole reason you need to do it yourself is because information does not get shared freely and widely. You practice, you learn, you know for the future.

> > So of course if we build a couple of naval ships every decade, maybe a giant train line every 30 years or similar, it will be done terribly.

> That's a reasonable explanation but it's not exclusive. You can still face these challenges and not do a terrible job. There is plenty of evidence this is the case.

No, there is no way to avoid doing a terrible job when you are out of practice. Maybe you can mitigate some of the worst issues and outperform expectations for someone out of practice, but you'll never be able to compete with those who do it regularly.

> > so far behind China in industrial capacity now.

> We just don't want to pay first world wages for the work.

No, even if the chinese were paid the exact same wages as americans they'ed be able to outperform because, having done this work for decades, they've gotten good at it. They've found new and innovative methods to make things more efficiently, they've invested in machinery and infrastructure to be more productive, and they've cultivated an industrial culture which makes it easier to learn and apply such lessons going forward.

China reunifies Taiwan: 20%

1.) too much corruption within the military. also no real war experience for 40 years

2.) not enough oil to supply all the ships needed for invasion. look at how Russia's column of tanks failed in the early invasion of Ukraine.

3.) China is broke and you need money for a war against US and Japan.

4.) China imports most of its food and oil

5.) Taiwan has very advanced anti-ship missile systems, homegrown and from US. and once a ship is sunk near the landing, that then prevents other ship from landing, basically piling up ship corpses.

US responds militarily: 70%

1.) Marco Rubio's first day on the job was to meet with AUKUS, which shows how important Asia and first island chain is

2.) Trump has said he would bomb China if China had occupied Taiwan under his presidency