the continued focus on cutting edge hardware is extremely foolish. the correct approach to computing in 2025 and beyond is to lock in hardware that's relatively easy to make, and write better software

the west is ignoring the resource we actually have, which is exceptional software engineers, in favor of a resource we don't have and can't get soon, which is TSMC

in general, asking old, giant, inflexible companies to solve problems is also a total waste of time and resources. they're dead players. we need new, small companies to be given the resources and flexibility to design and build new things across the board

none of these things will happen of course because the system is run by leeches. nobody in the west is actually trying to solve problems