Alvin Toffler Predicted This in 2006: Law at 1 MPH, Tech at 100

AI models and coding tools now ship at highway speed, while the laws meant to govern them are still stuck in first gear.

In their 2006 book Revolutionary Wealth, futurist Alvin Toffler and his wife and co-author Heidi Toffler compared society’s institutions to cars on a highway, each moving at a wildly different speed. Businesses and the knowledge economy, they wrote, race along at 100 mph, while the law and courts meant to keep pace crawl at 1 mph. They called this mismatch “desynchronization” and warned that the growing distance between institutions would eventually produce serious structural bottlenecks. Two decades later, anxiety over AI-driven job losses and the legal vacuum around generative AI prove that metaphor right in dramatic fashion.

The 100-to-1 Gap

The Tofflers’ speedometer sorts society’s major institutions into nine tiers. Businesses lead at 100 mph, followed by civil-society organizations at 90, families at 60, labor unions at 30, and government bureaucracies and regulators at 25. Schools trail at 10, international bodies at 5, political structures at 3, and the law comes in last at 1 mph.

The point isn’t that courts are lazy or legislatures move slowly on purpose. Law has to build precedent and social consensus, so caution is baked into how it works. The problem is what happens once technology grows explosively: the distance between the two widens out of control. When tech seizes markets and workplaces before any rules exist, it collides head-on with a legal system still catching up, and a large social vacuum opens up in between.

Where the Gap Is Breaking Down in AI Today

That gap is now causing real friction across the AI industry. Generative AI models train on hundreds of billions of tokens — a small piece of text, roughly part of a word, that AI models read and write — scraped from the open internet, and compress them into weights, but the fair-use doctrine — a US copyright exception that allows using protected work without permission for purposes like criticism or research — that courts rely on was built for a media world shaped by print and 20th-century broadcasting. The New York Times’ copyright suit against Microsoft and OpenAI, filed in federal court in December 2023, is still grinding through discovery and legal disputes as of 2026 without even a first-instance ruling. How long a final Supreme Court decision might take is anyone’s guess.

Developer tools face a similar standoff. In the GitHub Copilot class action, filed in November 2022, a federal court dismissed most of the plaintiffs’ claims under the DMCA — the US copyright law covering infringement takedowns and protections for digital works — in July 2024, ruling they hadn’t shown enough proof their code was copied verbatim without copyright management information (CMI) intact. But the breach-of-contract claims over open-source license violations have dragged on in court for years. Meanwhile, the Open Source Initiative (OSI), the body that defines what counts as open source, spent more than two years in internal debate before publishing its “Open Source AI Definition” (OSAID 1.0) in October 2024. Just agreeing on how much training data has to be disclosed to earn the open-source label took that long.

Government regulation shows the same lag. The EU’s AI Act — the European Union’s main AI law, adopted in 2024 — was first proposed in April 2021 and took more than three years to pass before taking effect that August. Its core rules on high-risk AI are still being phased in through 2026-2028 — by the time they’re fully enforced, the market will likely be full of autonomous agents nobody anticipated when the law was drafted.

The Dilemma: You Can’t Just Slow AI Down

As technology absorbs knowledge work faster than expected, fears about disappearing jobs are becoming real on the ground. A wave of autonomous agents that can code, analyze data and write documents faster than people has fueled anxiety across software engineering over whose role is next, along with debates over how roles should be restructured.

One proposed fix has been to deliberately slow AI’s pace of development. In March 2023, the Future of Life Institute published an open letter calling for an immediate, at-least-six-month pause on training AI models more powerful than GPT-4, and it gathered tens of thousands of signatures. But that push for restraint has gone nowhere in the face of global competition among great powers and the spread of open-source development. In a prisoner’s-dilemma dynamic, where a rival country or competitor keeps building even if you stop, hitting the brakes on technology is close to impossible in practice. Hoping the technology slows itself down isn’t a real answer.

Why the Law — Not Tech — Has to Find Its Speed

If technology can’t be slowed, the answer has to come from the other side. The institutions that hold the ultimate power to set and enforce society’s rules need to break out of their own outdated 1-mph stride.

While law sits comfortably in frameworks written years ago and keeps putting off decisions, workers and development teams on the ground carry a double burden. On one side, they live with the uncertainty of not knowing when their job might be replaced. On the other, they’re often left to defend themselves, with legal looking over their shoulder, over whether AI-generated code violates a license or whether an open-weight model — one where the maker is releasing model weights so anyone can download, run, or adapt them — runs into copyright trouble. The courts hold the power, but the people on the ground are paying for the uncertainty.

For law and regulation to keep that power while actually doing their job, they need to move away from a culture of lawsuits that take years to resolve after the fact, and build governance flexible and proactive enough to close the gap with technology. The future won’t wait, and a civilization racing at 100 mph can’t be protected by a legal system still moving at 1.

What to Watch

Watch for the first substantive ruling in the major AI copyright cases working through US federal courts, and whether lawmakers and courts elsewhere can actually break out of their 1-mph inertia and build regulatory frameworks flexible enough to match the pace of technological change.