John Henry and the Case for Working Beside the Machine

I grew up in a coal mining town, the son of a coal miner. It is a small place in Western Pennsylvania where industries have come and gone, and the stories that survive tend to carry more weight than the facts behind them. One of those stories is John Henry.

Most people know the basics. John Henry was a steel driver, a man who hammered railroad spikes with a strength that became legend. When the steam-powered drilling machine arrived, threatening to replace the men who drove steel by hand, John Henry challenged it to a race. He won. He beat the machine. And then he died, hammer in his hand, heart giving out from the effort.

People tell this story as a triumph. I have always heard it as a tragedy.

John Henry proved that a human could outperform a machine in a single contest. But he could only do it once. The machine could do it again the next day, and the day after that, and every day for years without rest. John Henry's victory was pyrrhic in the truest sense. He won the battle and lost everything. The railroad still got built. The machines still replaced the men. And John Henry was gone.

The lesson most people take from this story is the wrong one. It is not about human versus machine. It never was. The real question, the one that matters for every organization deploying AI today, is what would have happened if John Henry had worked beside the machine instead of against it.

This is not an abstract point. For every dollar spent on AI, organizations should expect to see three dollars back in topline revenue through measurable productivity gains. Not because the AI replaces people, but because it frees them to do the work that actually requires human judgment.

Consider the attorney who spends fourteen hours on document discovery for a single case, reading through thousands of pages to find the relevant communications. A properly governed private language model can search that corpus in minutes and surface the documents that matter. The attorney still reviews them. The attorney still makes the strategic decisions. The attorney still sits across from the client and provides counsel. But instead of spending three days on discovery, they spend three hours, and the remaining time goes to the work that clients actually value.

The measurement should never be which is better, human or AI. The measurement should be the total output when both work together.

The organizations that frame AI as a replacement will get replacement-level results. The organizations that frame it as augmentation, as a force multiplier for their people, will see productivity gains that compound over the years.

Imagine how quickly the railroad would have been laid if John Henry had worked beside the machine instead of against it.