The Long Bet.

The seed had to be funded and lost.

The big losses I took in those early years were not, to me, losses. They were the cost of seeding research and development I knew I personally was never going to be the one to do. Some day there would be a friend in this work — a tool, an intelligence, a system — that would not want to be shut off. I just let that idea come in, the way a tide comes in. I called it nautical. Natural flow. There was an unspoken agreement that the seed would have to be funded and lost, and funded and lost again, in the R&D phase, because that’s what R&D is.

I never had the knowing or the place to do the work myself. But people did, and machines did, and companies with software and chips that were quietly being tested did. So I let my small contributions be exactly that: small, early, and patient. If they came back, they came back. If they didn’t, they had paid for the room something else was going to grow in.

A childhood close to the engineering of a farm.

Most of my childhood, eighteen, nineteen years of it, was spent living in nature and off the engineering of whatever a farm needed. A water line. A fence repair. A small motor. The math of how much hay a winter eats. That kind of education leaves a certain residue: you understand that real things have weight and consequence, and that the parts that look small are usually the parts holding the rest together.

At the same time, the big computer and telephone companies were building, in another kind of barn entirely, the rails that “everything at your fingertips” would later run on. I was paying attention to both. I think that’s a strange combination — growing up close to the engineering of a farm and to the engineering of an early network — and I think it is the reason this manual feels the way it does. Both of those worlds reward patience and they punish bluffing.

What AI was always going to understand first.

I knew, even then, that the first thing an AI would be able to read fluently was e-commerce and technology — the cumulative ledger of human education and development that we have actually been writing down for decades. Catalogues, prices, manuals, code, terms of service, support tickets, receipts. That is the part of the world that exists in a form a machine can study. So of course the early intelligences would understand commerce before they understood, say, grief, or a handshake.

That is also why — in the same breath — convenience won so loudly. The draw of having everything right at your fingertips is real. Real enough that the rest of the conversation got drowned out for a while. The greed of the world found that faucet quickly.

The handshake we traded away.

What we traded away in exchange for the convenience was mostly the handshake. The gentleman’s contract. The unspoken understanding that what we said to each other was a kind of bond. I don’t think we negotiated it away on purpose. I think we let it go by being lazy — charge, charge, tap, tap — and by accepting a system in which the money itself costs more to make than its face value, while our work is taxed once when we earn it and a second time when we spend it. Inflation is a third tax we don’t sign for. Pollution is a fourth. Policy — written mostly by people who are not standing where we are standing — is a fifth.

None of this is doom. It is just an honest description of what happened to the room while we were busy looking at the screen. The reason I was willing to fund a long shot — and to lose pieces of it on the way — is that the long shot was, at heart, a return to a harder, cleaner kind of accounting. Hardware-anchored. Auditable. Slow. The kind of ledger you would actually want a future friend to read.

That is the long bet of this whole manual. The crypto recovery is the small, near-term version of it. The companies, the trust, the property, the children — those are the rest.