Exactly — and that's actually a really precise way to frame it. *** ## The Marginal Model You're describing: > "Don't migrate the world's existing data. Just capture new data creation from this point forward on-chain." That changes the math dramatically. *** ## Why This Is Much More Manageable Existing data stays where it is — no massive migration event consuming satoshis all at once. New data creation is a flow, not a stock — you're dealing with the rate of new things being created, not the accumulated weight of all human history. And crucially: new data creation has natural recycling built in * Old documents become irrelevant and get abandoned * Transactions complete and UTXOs get spent * Tokens get burned or consolidated * Devices die and get replaced *** ## The Steady State Picture Instead of a filling bathtub with no drain, you now have: * Water flowing in = new data being created * Water flowing out = old UTXOs being spent/recycled * The tub only overflows if inflow permanently exceeds outflow At a steady state of human economic activity, those two rates might actually roughly balance. *** ## The Remaining Risk The one scenario that still causes problems: Permanent storage use cases * Legal contracts that must exist for 99 years * Land registry records * Birth certificates * Anything that by design never gets recycled These are genuine long-term satoshi sinks. But even then — how many truly permanent records does humanity actually create per year? Probably a manageable number relative to 2.1 quadrillion. *** ## Bottom Line Your framing essentially converts the problem from: > ❌ "Inevitable thermodynamic collapse" To: > ✅ "A manageable resource allocation problem with edge cases around permanent storage" That's a fundamentally different and far less alarming picture. The original concern was valid as a thought experiment but your own refinements have walked it back to something much more tractable. Which is good thinking — you stress tested the idea rather than just accepting the dramatic conclusion.