I watched a project die last spring. Not with a crash, not a hack, nothing dramatic. It just went quiet. The token that had been the whole point of it slid down, the rewards thinned out, and one Tuesday the people who had been posting hourly numbers simply stopped showing up. The Discord went from 4,000 messages a day to a pinned announcement nobody read. What struck me wasn't that it failed. It's that the moment the payout stopped, there was nothing left. No product. No habit. No reason to open the app. The reward was the thing. And when the reward left, the thing left with it.
That's been rattling around in my head ever since, because it points at a test I now apply to almost everything I look at.
Would you still do this if you never got paid for it?
I don't mean that as a moral scold. I mean it as a filter. Most short-horizon designs fail the test instantly. A yield farm that exists to farm its own yield is worth exactly zero when the emissions dry up. A game whose only verb is "claim" is not a game, it's a faucet with a loading screen. The people playing it aren't players, they're extractors, and extractors leave the second the vein runs dry. You can dress it up with pixel art and a roadmap, but the underlying structure is a countdown timer disguised as a product.
Compare that to the things that lasted.
Bitcoin's block reward halves roughly every four years, and it has done this four times since 2012 — 50 coins per block in the beginning, then 25 in November 2012, then 12.5 in 2016, then 6.25 in 2020, and after April 2024, 3.125. Every halving, the loudest prediction is that miners will capitulate and the network will spiral. Every halving, the network keeps producing blocks roughly every ten minutes anyway. Why? Because a meaningful chunk of the reason people run and use that system was never purely the subsidy. It's the settlement. It's the property of being able to hold something nobody can print more of. The financial incentive is real and it matters, but it sits on top of a thing that would have a reason to exist even if you shrank the reward. That's the difference between a foundation and a sugar high.
I think about the open-source world the same way. Linux has no token. Wikipedia has no token. SQLite, which quietly runs inside your phone and your browser and probably three apps you have open right now, has no token and its author has said he intends to keep it in the public domain. These things compound for decades precisely because they were worth doing without a payout. The contribution was the point. And here's the part nobody wants to hear during a bull market: those unglamorous, unpaid-in-spirit projects have generated more durable value than 99% of the things that launched with a countdown and a claim button.
What a decade selects for
When you extend your time horizon, the whole scoring system flips.
On a one-hour horizon, the winning move is to be loud, be first, and be gone before the music stops. Attention is the asset. On a ten-year horizon, attention is nearly worthless — it evaporates faster than anything. What survives a decade is stuff that is quietly useful whether or not anyone is watching. Habits. Trust. Code that keeps working. A community that would still gather if you deleted the leaderboard.
Compounding only works on things that persist. That sounds obvious until you notice how many designs are structurally incapable of persisting. You cannot compound a faucet. The math of compounding assumes the base survives to the next period, and a payout-first system is engineered so the base leaves the moment the payout does. It's like planting a tree that's programmed to uproot itself the day you stop watering. You will never get shade from it, no matter how much you water it early.
Here's what nobody tells you about patience: it isn't passive. Waiting isn't the same as doing nothing. The patient move is to build or hold the thing that has intrinsic value — value it would still have if the market closed for five years — and then let time do the boring work of accumulation. That's active. It requires you to say no to the loud thing over and over while everyone around you is getting rich on paper this quarter. That's the tax patience charges, and most people can't pay it.
I got this wrong myself for years, by the way. I used to conflate "early" with "smart." Being early to a payout felt like insight. It was mostly luck with good timing, and I mistook the two, which meant I couldn't repeat it and didn't understand why. What actually repeats is picking things with a reason to exist and staying long enough for the reason to matter. That's not a trade. It's closer to gardening.
The test, made concrete
So when I look at anything now — a chain, a game, a community, a piece of software, a habit — I run the same three questions:
- Strip out the reward entirely. Is there still a reason for this to exist? If the honest answer is no, I'm looking at a countdown timer.
- Who's here for the thing versus here for the payout? Watch what happens on the boring days, not the launch day. The boring days tell the truth.
- Does this compound, or does it only extract? A system that produces something — trust, utility, a working product, a real habit — can compound. A system that only redistributes tokens from newcomers to insiders cannot, no matter how the chart looks this month.
None of this is a prediction. I'm not telling you what any price will do, and I'd be lying if I pretended to know. Timeframes are the only thing I trust myself to reason about. And on a long enough timeframe, the payout-first stuff has a way of returning to its intrinsic value, which is frequently zero, while the things that were worth doing anyway keep quietly stacking blocks whether the crowd is watching or not.
Everyone wants the next hour. I want the next decade. Not because the next decade pays more — though it usually does — but because the decade-long question forces you to hold only what deserves to survive it.
The games worth playing are the ones you'd still play with the prize turned off. Build those. Hold those. Ignore almost everything else.