The Industrial Seed Crystal: What Modern Survivors Would Actually Need to Cache
If survivors keep their modern knowledge, the cache problem inverts: you no longer need an instruction manual, you need the physical bottlenecks that take centuries to recreate by hand. Priority order is the electricity loop, machine tools, refined raw materials, finished semiconductors, and dating instruments.
The classic cache problem assumes survivors forgot everything. But if they retain modern knowledge, the bottleneck is not knowing how a transistor works, it is having refined silicon, a vacuum chamber, and photolithography to actually make one. The cache strategy flips from 'instruction manual plus microscope' to prepositioning the *industrial seed crystal*: the specific objects that would otherwise take centuries to recreate from hand tools. **1. Close the electricity loop on day one.** Modern humans without power are functionally medieval. Electricity has a chicken-and-egg problem (you want electrolysis to refine copper, electric machine tools to make generator parts, furnaces to make magnets), so the top cache priority is breaking that loop immediately: industrial generators sealed in inert gas, kilometers of insulated copper wire, neodymium magnets (which hold magnetization for millennia below their Curie point if shielded from corrosion), plus pre-made bearings, brushes, and commutators. A waterwheel or wind frame is trivial to build by hand once you have a generator head. **2. Machine tools, the tools that make tools.** A lathe is what lets you make another lathe. Cache a complete small machine shop (precision lathe, mill, drill press, surface grinder) plus the irreplaceable accessories: collets, chucks, taps and dies in every standard thread, carbide cutters, and crucially the precision metrology that took a century to develop, including a set of gauge blocks, granite surface plates, micrometers, and sine bars. Without gauge blocks you cannot make accurate gauge blocks; with one reference set, modern machinists replicate everything. Packed in cosmoline or sealed in nitrogen, they survive indefinitely (see Gauge Blocks: The Reference Standard That Bootstraps Machine-Shop Precision). **3. Raw materials that hide entire supply chains.** Hand-tool civilization can smelt iron and make bronze but cannot practically produce aluminum (needs cheap electricity and cryolite via Hall-Heroult, see The Hall-Heroult Process: How Electrolysis Made Aluminum Affordable), titanium (Kroll process), pure tungsten, rare earths, or semiconductor-grade silicon. Stockpile ingots of each, plus catalysts that are annoying to make from scratch such as vanadium pentoxide for contact-process sulfuric acid. **4. The things you genuinely cannot remake quickly: semiconductors.** Even with full knowledge, building a fab is realistically a 30-50 year project because photolithography needs ultra-pure optics, clean rooms, plasma etchers, and deep-UV sources, each with its own supply chain. So cache *finished* chips in inert atmosphere (microcontrollers, op-amps, MOSFETs, regulators, ADCs) plus passives by the bin. A few meters of rock shielding protects them from cosmic rays. You will not make new ones for a long time, but you can bootstrap instrumentation and control systems. **5. The non-obvious essentials people forget.** A way to know *what year it is* on waking (a long-half-life radioactive sample of known initial mass, or a precision astronomical reference such as a stone observatory aligned for the petrification year, so survivors can measure precession), and redundant engraved metal *maps to the other caches*, because losing cache locations makes the cache worthless. Realistic timeline with all this prepositioned: survivors reach roughly 1950s technology within a year, 1990s within five to ten, and stay below 2020s until they rebuild a fab, the great patient project that is century-scale even with full knowledge. That is still an order of magnitude faster than starting cold.