Innovation is often described as if it can be planned, taught, managed, or scaled through a clean formula. Eugene Fitzgerald, the Merton C. Flemings SMA Professor in MIT’s Department of Materials Science and Engineering, argues that this misses the deeper reality.
In his latest book, The Invisible Engine: Why Innovation Evades Control, Fitzgerald frames innovation as a process that moves across science, economics, industry, and society. The central question is not only how new knowledge is created, but how it becomes useful in the world.
Innovation Is More Than a Breakthrough
Fitzgerald’s view comes from the last 10 years of work in research-to-market activity, including the MIT and Masdar Institute Cooperative Program and the MIT-Singapore Alliance for Research and Technology. Those large research efforts did not fit neatly inside one discipline.
Scientific journals can capture discoveries within specific fields. But Fitzgerald saw a gap around larger projects where research, economic forces, industry needs, and social context meet. His book grew from an effort to connect those pieces and explain how innovation works at that scale.
That distinction matters because discovery and innovation are not the same thing. A discovery may be important within physics, materials science, or another field. Innovation requires more: it must move through real institutions, real markets, and real constraints before it creates value.
The Invisible Engine Is Decentralized
Fitzgerald describes the “invisible engine” as a decentralized collective intelligence. It is made up of people and companies whose separate actions can eventually produce surprise in the marketplace.
The idea of surprise is tied in the source to Frank Knight, an economist from the early 1900s who studied the Industrial Revolution around him. Fitzgerald explains that the entrepreneur takes on uncertainty by bringing something into the world without knowing exactly what will happen.
That uncertainty is not a flaw in the system. It is part of the mechanism. The entrepreneur acts before the outcome is clear, and the reward comes when others are surprised that the product is wanted and that it can be made.
This makes innovation difficult to control from the top down. It is not simply a management exercise or a pipeline from lab to market. It is a changing process involving many actors, each responding to possibilities that are not fully known in advance.
Strained Silicon Shows the Path Is Not Linear
Fitzgerald’s experience with semiconductor technology shaped the book’s argument. At AT&T Bell Laboratories in the 1990s, he and a colleague discovered a way to strain silicon in thin-film form with very few defects. From a physics perspective, the result was significant.
But Fitzgerald wanted impact beyond scientific recognition. When he asked his manager what should happen next, the answer was to speak with the marketing people at AT&T. That response highlighted a central challenge: even strong industrial labs could create important things without always being able to commercialize them.
After Bell Labs, Fitzgerald came to MIT, where he says he could keep uncertainty open across different elements and look for convergence in multiple directions. He later started a company. The path eventually included a settlement with Intel over a patent dispute after the industry found that strained silicon was needed to extend Moore’s Law.
Moore’s Law is described in the source as the semiconductor industry’s long-standing trend of increasing the number of transistors on chips roughly every two years. Fitzgerald’s point is that the route from lab discovery to industry adoption was not tidy. It moved through Bell Labs, MIT, a startup, and then broader use.
Three Kinds of Research Investment
One misconception Fitzgerald identifies is the assumption that all research investment works the same way when the goal is economic impact. He separates research investment into three categories, each with a different purpose.
- Altruistic science is investment in academic institutions to produce educated people. Fitzgerald says it is not done in the context where ideas eventually need to succeed, and that its direct economic yield is “basically zero.”
- Strategic research is organized around a goal rather than a single scientific or technical area. Fitzgerald gives the example of a new F-35, where the customer is the government and several fields may need to advance together.
- Fundamental innovation covers the process from research to economic growth, even across 10-, 15-, or 20-year time horizons.
Fundamental innovation is the category where Fitzgerald places the full complexity of the work. It involves technology, implementation, and market at the same time. Technology asks what is physically possible. Implementation asks how something can be built and delivered. Market asks who will adopt it, and why.
That framing changes the meaning of innovation. It is not just a bright idea waiting to be funded. It is a continuing effort to test what is possible, how it could be delivered, and whether the world will value it.
Why This Matters Now
The book arrives as artificial intelligence is reshaping how people think about knowledge, research, and innovation. Fitzgerald’s argument points to a broader issue: society cannot invest well in the future if it misunderstands what innovation requires.
He wrote the book for individual innovators and students; researchers and faculty; corporate leaders; research funders; and policy-makers. The common thread is that each group has a stake in decisions about careers, institutions, public investment, private investment, and the far future.
The clearest takeaway is that innovation cannot be reduced to a single invention or a single institution. It happens when knowledge, markets, implementation, and uncertainty remain in contact long enough for value to emerge.