Three Disruptions

This chapter was drafted before the Foundation adopted Synthetic Intelligence (SI) as its preferred term. The vocabulary has been preserved as written; the reasoning behind the rename lives on the Mission page.

Every two or three generations, a technology arrives that does not just change one industry but rearranges the structure of ordinary life. It is rare. Most of what gets called innovation is a matter of better packaging or shorter cycles. The genuine reshufflings, the ones that (looking back from a century later) divide the world into a before and an after, happen perhaps three or four times in a century. We are inside one now. To understand what the Webspinner Movement is for, it helps to look at the last three.

The railroads

In 1830, the typical American moved at the speed of a horse. By 1870, a person with a ticket and a sandwich could cross the continent in eight days. The railroads did not just shorten travel time. They created the modern timezone (local sun-time was unworkable once trains had to keep schedules) and the modern corporation, since no entity smaller than a corporation could finance and operate a railroad. They also created a particular American figure: the railroad baron. Vanderbilt, Gould, Hill, Stanford, Huntington. These were men who built things, and the things they built carried wheat and beef and immigrants and mail across a continent that had previously been crossed by oxcart. They were not, on balance, gentle men. The land grants were enormous. The labor practices were brutal. The political corruption was open. Frank Norris wrote a novel about it called The Octopus, in which the Southern Pacific Railroad sat in California like a great suctioning creature, its tentacles reaching into every legislature and every wheat field. The novel was a bestseller. It changed approximately nothing.

What changed it, eventually, was a combination of three things: trust-busting (Theodore Roosevelt and Sherman Antitrust), labor organization (the railway brotherhoods, the strikes of 1877 and 1894), and time, which turned the original barons into their grandchildren, who sometimes funded libraries. The railroads remained essential. The railroads remain essential. But the political settlement that followed the Gilded Age (the recognition that a transformative technology cannot be left entirely in the hands of its first owners) is what allowed the country to absorb the railroads without being eaten by them.

Electricity

In 1882, Thomas Edison opened the Pearl Street Station in lower Manhattan. By 1925, electricity in major American cities was an unremarkable utility. You flipped a switch and it worked. By the same year, only about ten percent of American farms had electrical service. The technology had been around for forty years and had transformed urban life entirely; rural America was still trimming oil lamps. The reason was simple. Private utilities had no commercial interest in stringing wire down a country road that served eight farms. The economics did not pencil. The technology existed; the access did not.

What changed it was the Rural Electrification Administration, signed into existence by Franklin Roosevelt in 1936. Within a decade, ninety percent of American farms had electricity. The mechanism was not state ownership of the power industry (Edison and Westinghouse and the local utilities went on operating exactly as they had) but a federal subsidy and the formation of farmer-owned electrical cooperatives that strung the wire that the private utilities would not. The cooperative model, mostly invisible to people who have never lived in the rural places it served, is one of the great institutional inventions of American history. It looked nothing like nationalization. It looked nothing like the trust structure of the great utilities. It was a third thing: owned by its users, governed by its users, designed to serve communities that were uneconomic to serve under the existing arrangements.

You can see, if you squint, where this is going.

Internal combustion

The Model T put a private automobile in the driveway of an American household that had, fifteen years before, never imagined owning one. It transformed the geography of the country, the dating practices of every adolescent over five feet tall, the supply chains of every business, the location of every grocery store. It also locked the entire civilization into a fossil-fuel dependency that is only now, with great difficulty, being unwound. Henry Ford was not, by any honest reading, a great moral figure. The automobile he gave the country was a real democratization, and the cost of that democratization is being paid in glaciers.

The lesson of the internal combustion engine is the most painful of the three. The technology was genuinely emancipatory. It also locked in a path dependence that the world is now spending trillions of dollars trying to redirect. The decisions made in the first decades of automotive adoption (the road network instead of streetcar networks, the suburbs instead of denser cities, the gasoline tax instead of carbon tax) were made by engineers and politicians who had no way of knowing, in 1925, what the carbon load on the atmosphere would be in 2025. The general principle is straightforward: the structure you build into a transformative technology in its first decades is the structure you live with for the next century. You do not get to retrofit the foundations.

What the three have in common

All three were genuine. They actually changed ordinary life. They were not hype.

All three concentrated wealth, in the first generation, into a small number of hands. The barons, the utility holding companies, the automobile manufacturers and oil firms. None of this was illegal at the time. It was simply the natural shape of an unregulated transformative technology in its first decades.

All three required, eventually, some form of public counter-institution to balance the private one. The interstate commerce commission for the railroads. The rural electrification cooperatives for power. The fuel taxes and the highway trust fund for the automobile. None of these were socialism. All of them were, in their context, recognized as necessary.

And in all three cases, the public counter-institution arrived late. By the time it arrived, the structural shape of the technology’s deployment was already mostly set. The interstate commerce commission could regulate the railroads, but it could not unbuild the company towns or unhand the land grants. The REA could electrify the farms, but the broad shape of utility ownership had been determined fifty years earlier. The fuel taxes could discourage gasoline use at the margin, but they could not undo the suburban geography that gasoline had already produced.

The work of the public counter-institution, in every case, would have been ten times easier if it had been started ten years earlier. Most of the great mistakes of all three eras came from waiting.

The fourth

We are now inside a fourth disruption, and it has the same shape. Artificial intelligence is real. It is not hype. It is going to rearrange ordinary life in ways that, ten years from now, will seem as inevitable as electric lights and as unremarkable as a car in the driveway. The economic projections, which we cite at length in the business plan, point in only one direction. The consumer AI market is on a path from ninety-two billion dollars in 2024 to six hundred and seventy-four billion by 2030. This is not going to slow down. It is not going to be banned by anyone in any country. The question is not whether AI gets built; the question is who owns the AI tools that ordinary people use, and how those tools are structured.

Right now, the answer is what you would expect from history. A small number of well-capitalized firms (OpenAI, Anthropic, Google, Meta, Microsoft) are building the foundational infrastructure. Their products are extraordinary. Many of the people working at them are sincere. We are not going to demonize them. They are doing what their incorporation papers obligate them to do, which is to maximize value for their shareholders. They will build excellent tools, and they will charge for those tools, and the tools will end up most available to the people who can most afford them. This is not corruption. It is the foreseeable shape of a transformative technology in its first decade under a particular kind of corporate structure.

What the moment calls for is the same thing every previous moment has called for: a public counter-institution. A version of the technology that is built for, owned by, and accountable to the people who will use it, not to the shareholders of a public company. A REA for AI. A Wikipedia for AI. A Patagonia for AI.

It does not have to be the only version. The commercial AI labs can keep doing exactly what they are doing. We need them to. They are pushing the technical frontier. The Webspinner Foundation is not in competition with them; it sits beside them, the way Wikipedia sits beside the Encyclopedia Britannica, the way Linux sits beside Windows, the way the rural electric cooperatives sat beside Consolidated Edison. A pluralism of structures, with at least one of those structures owned by its users and locked to its mission.

The lesson of the previous three disruptions is that this counter-institution always eventually gets built. The lesson is also that it always arrives a decade late. We are trying, this time, to arrive on time.

That is the work. That is what Webspinner is for.