Like any good American boy raised on the right scifi, I spend some time pondering the origins of life on Earth. The model I find most plausible is that abiogenesis - the creation of life from non-living precursor materials - is actually fairly likely over geologic timescales on rocky planets in the habitable zone, but the presence of life itself makes it impossible after the first real success. Suppose the recipe is fairly simple. Let's assume it happened in tide pools, and there were under ten critical precursors: The right types of rocks on the beach, a dilute suspension of 2 or 3 nonpolar molecules flowing onto the beach via one river which originates in volcanic geology, some clay to provide nice high-surface-area substrates for the action, and a nice concentration of simple organics (read: proto-life food) like HCN and HCHO created photochemically in the atmosphere, rained into a sedimentary rock deposit over millenia, and then washed onto our lucky beach by a second river which meets the volcanic one a few kilometers upstream.
Suppose that's all it took, and if you leave a beach like that to do its thing for 50 million years you get primitive life. What happens next? The proto-life is going to eat all the primitive organics on that beach. It might wipe out the geologic supply before it lucks into any other viable energy sources and die out, but since we exist we know at least one lineage made it past that filter and expanded into the ocean. Then what? In perhaps just a few thousand years the proto-life fills the oceans and eats all the primitive organics on every beach. As a thoughtless greedy evolving system, it razes its own cradle.
This is a frustrating system to study or theorize about because something "easy" for an entire planet to luck into over millions of years can be utterly impossible for human researchers to recreate within a lifetime, or even the lifetime of civilization so far. If the odds are one in ten trillion per liter of optimal-tidepool-water per year, and the pools can only be 6 inches deep for wet-dry cycle reasons, humans would need to cover all of Rhode Island in tidepools for 20 years. Tidepools which we'd need to scrupulously sterilize of all modern mainline-life to protect the proto-life experiments from competition, and spam with advanced sensors to actually catch the protolife when it shows up. Young earth on the other hand maybe1 had ten million liters of viable pools at any given time and they were all sterile because life didn't exist here yet, so with those odds it would get a shot every million years. Current estimates say Earth got it done ~300-900 million years after the oceans condensed, so a few hundred false starts and then our lucky progenitors made it off the beach sounds about right.
I think a lot of complex systems with resource constraints are like this. Any time you have more than a couple necessary precursors to create a thing, and the created thing consumes at least one of its precursors greedily and globally, that thing will be unique.
The thing on my mind is Shenzhen, or rather the whole Chinese manufacturing behemoth we metonym as "Shenzhen". A nascent manufacturing ecosystem needs smart, hardworking, underutilized young workers, a big pile of demand for easy to build products at high prices, risk capital and ambitious leaders to deploy it, lots of cheap land, a low drag regulatory environment, affordable transportation rails within the ecosystem and out to the demand, etc. Once a mature ecosystem like Shenzhen exists, it fulfills all the demand for easy products and crashes their prices. Its workers and firms gain experience, develop advanced manufacturing techniques, and build business processes to bang out new high tech products at incredible speed with shockingly low costs.
Shenzhen got to start with low end textiles, furniture, toys, and basic electronics, but today they ship all those products to every shore. The risk capital pouring into Shenzhen grew exponentially over the years, but each tranche paid itself off by finding or creating new demand to satisfy. Any given year looks like they're investing far more than they're making, but if you track the cohorts right it's screamingly profitable the whole time and they just kept doubling down. They rode a natural growth curve, and now if we want our own Shenzhen the path they took is closed to us. Even if America had every other precursor to create our Shenzhen, the global supply of demand is being consumed by Shenzhen 1.0.
This isn't as simple as burning a few trillion and a proverbial Rhode Island to fix, either. You need millions of workers employed and competitively motivated to bootstrap a few hundred thousand of them into highly skilled experts who can push the frontier and compete with Shenzhen. Stubbing out the hormetic strain of profitably serving real market demand is hard - if you just throw dollars at manufacturing for its own sake you get an incompetent Potemkin workforce.
The least expensive2 path out I can see is for the government to create enormous demand and bar Chinese components from being used to satisfy it, but that requires the government to exercise good taste on questions of engineering, streamline procurement, and resist the urge to award work based on political goals other than 'build great things' and 'win future wars'. I don't think that's going to happen, so unfortunately I think our best bet is that Elon Musk's conglomerates bootstrap a workforce for us by sheer power of will and his unique access to capital, and that somehow compounds fast and hard enough to get us back in the race. Maybe if we solve AI and autonomous robotics before China it can happen.