If you’re feeling anxious about AI replacing your role, you’re in good historical company. Seth Seter’s recent essay, 'How to Be a Good Farrier,' draws a sharp parallel between the rise of the automobile in the early 20th century and the current AI revolution. The core message is comforting but demanding: technology doesn’t wait for us to be ready, but those who master their craft and adapt will survive. It’s not about hating the machine; it’s about being so good at your job that the machine can’t replace you.

The Car Was a 'Fad' Until It Wasn't

In 1900, New York City relied on 130,000 horses, creating a massive economy around blacksmiths, teamsters, and stable hands. Early cars were seen as dangerous toys for the rich, with the New York Times calling them 'uncanny' and 'indecent.' Critics argued they were unsafe, too expensive, and a threat to public order. Sound familiar? Today, we hear similar arguments about AI being a 'blurry JPEG of the web' or a threat to creativity. But while the opposition was loud, it was powerless against the economic reality of the Model T, which dropped in price from $825 to $260 between 1909 and 1925.

What Happened to the Farriers?

The transition wasn’t pretty for everyone. Census data shows that between 1910 and 1930, the number of teamsters and carriage drivers plummeted from 443,845 to 111,224. The number of hostlers and stable hands dropped from 63,388 to 6,654. However, this wasn’t a total extinction event for the trade. The American Farriers Journal estimates there are still about 27,500 farriers in the US today. The ones who survived were the best in their field, working on expensive horses where quality mattered more than speed. The mediocre farriers, who relied on the sheer volume of horse traffic, were the first to disappear.

The Modern 'Canary in the Coal Mine'

We are seeing this pattern repeat now. Stanford’s Digital Economy Lab found that workers aged 22 to 25 in AI-exposed jobs are employed 19% below where they would be without AI, while experienced workers show no such gap. Indeed reports that 69.3% of software job postings in early 2026 were for senior roles, with entry-level postings at just 4.5%. The work that is disappearing first is the 'boilerplate'—the tasks that didn’t require deep judgment. If your job is just closing tickets or writing basic code, you are in the danger zone.

Key Takeaways

  • Master the Craft, Not Just the Tool: Know why the system is shaped the way it is. You can’t tell if AI helped you unless you could have done the work yourself.
  • Own the Outcome: Anyone can generate code, but few will stand behind it under load. Be the person who takes responsibility for the result.
  • Move Toward the Replacement: Like the blacksmiths who started repairing car axles, identify where your skills transfer to the new industry.
  • Stop Waiting for Bans: Regulatory attempts to stop technology, like Vermont’s 1894 law requiring a person to walk ahead of steam vehicles, always fail.

The Bottom Line

Don’t bet on the horse coming back. Be a good farrier by mastering your domain, using AI to enhance your judgment, and owning the final product. The mediocre will be automated; the excellent will be augmented.