Electricity Took Decades to Pay Off. Here Is What That Teaches Us About AI.

Electricity is the cleanest case of a technology whose payoff arrived long after the technology did. When electric power first reached factories, the transformative effect was not immediate. Many plants simply swapped electric motors in for steam engines inside their existing layouts. Operations got a little cleaner and a little more flexible, but the full productivity potential of electrification stayed locked up.

The deeper gains came later, when factories were redesigned around what distributed power actually made possible: machines placed by the logic of the work rather than by proximity to a driveshaft, more flexible layouts, new production methods, and different management practices. The motor was available for years before the reorganization caught up.

The point is not subtle, but it is easy to forget under pressure. The electric motor was not enough. The organization of production had to change, and that change required experimentation, capital, managerial imagination, and time.

A steam-powered factory was built around a central power source. Belts and shafts transmitted mechanical power throughout the building, so machines had to be arranged to reach the transmission, not to suit the flow of work. The architecture of the factory encoded the constraints of its power source.

Distributed electric power dissolved that constraint. Each machine could have its own motor, so machines could sit wherever the work made sense, production could flow differently, and labor could be reorganized around the process instead of the driveshaft. But unlocking that required tearing up assumptions that had been baked into the building, the equipment, and the management routines.

So firms had to learn what the new technology made possible, and learning is slow. The first instinct, replicate the old layout with a cleaner power source, was rational and incremental. It just left most of the value on the table.

The same lag reappeared with information and communications technology. The Chicago Fed put it plainly: realizing the benefits of ICT required substantial complementary investments in learning, reorganization, and the like, so the measurable payoff could be long delayed, with electric power offered as the historical analogy.

That reframes the productivity delay. It was never evidence that the technology lacked value. It was evidence that organizations had not yet built the complementary systems required to capture the value. The lag lives in the organization, not in the tool.

This is the single most useful thing electricity teaches us about AI. A gap between adoption and measurable productivity is normal. It is the expected shape of a general-purpose technology working its way through institutions that have to change to accommodate it.

Companies should plan for a visible gap between AI access and AI productivity. Giving employees AI tools will improve individual tasks quickly: drafting emails, summarizing meetings, generating code snippets, producing first-pass research, rewriting customer responses. Those wins are real and they arrive fast.

Enterprise-level value is a different animal. It requires workflow redesign, new review mechanisms, changed roles, updated controls, revised performance metrics, and new forms of accountability. None of that comes from the tool. All of it comes from the organization choosing to rebuild the work.

A sales team that merely lets representatives use AI to write emails will get modest gains. A sales team that redesigns account planning, call preparation, CRM hygiene, follow-up generation, objection handling, and coaching around AI is operating in a different class of improvement. Same tool, different result, because one of them rebuilt the system.

The common mistake is to treat the transition as install the technology and wait. History says that is wrong. The better model is install, redesign, train, govern, measure, and iterate. AI will produce quick wins, but its durable advantage comes from redesigning processes around AI-enabled work.

That distinction has budget and patience implications. If leadership expects transformation from access alone, the pilots will disappoint and the program will lose support right before the complementary investments would have paid off. If leadership expects to fund redesign, the early unevenness reads as exactly what it is: the cost of learning.

Electrification rewarded the firms that reimagined the factory, not the ones that waited for the motor to do it for them. AI will reward the firms that reimagine the workflow, not the ones that wait for the model to do it for them.

Electricity also teaches patience without passivity. Organizations should not conclude that AI has failed because early pilots produce uneven returns. Nor should they coast on vague optimism that it will all work out. Both responses abdicate the actual work.

The disciplined response to early unevenness is structured learning. Establish baselines. Test specific workflows. Identify where AI helps and where it fails. Redesign the work. Then repeat. Productivity emerges from the system built around the tool, and systems are built deliberately.

If you remember one thing from the electricity story, make it this: the technology sets a ceiling, but the organization decides how close you get to it. The motor made a better factory possible. People had to go build it.