Last year, I received a call from the National Science Foundation, inviting me to teach entrepreneurship to a group of scientists in a new program called the Innovation Corps. The NSF, of course, is the government entity that funds science and engineering research. The agency has an annual budget of $7 billion but has never been particularly adept at identifying and commercializing projects.

The I-Corps, I learned, was an attempt to remedy that. The year-old program awards NSF-funded research teams $50,000 and ongoing, intensive mentoring from entrepreneurs, venture capitalists, and tech developers. I agreed to participate but was skeptical. Successful entrepreneurs--especially those in Silicon Valley, where I have lived, worked, and taught for the past 35 years--move fast. Could a group of scientists, who tend to operate at a far more methodical pace, ever really get it?

As it turned out, the teams already had the most important things they needed to get started: their research and some ideas about how it might be commercialized. I simply asked them to test those assumptions by talking to as many potential customers as they could. In my Stanford M.B.A. classes, I call this customer development, but the lab rats immediately recognized it as something else: the scientific method in action.

Still, the first few weeks weren't easy. Remember, these were scientists who are most comfortable in the confines of their labs. Now they were being asked to make cold calls and talk to people out in the world about what they needed and how much they were willing to pay for it. Meanwhile, investors, entrepreneurs, and I talked to them about things like finding sustainable business models and offered encouragement--and skepticism--when it was needed.

One team from Yale, called Red Ox, was working on electrochemical technology to desalinate brine, the salty waste­water that results from a number of industrial processes. The researchers came into the program thinking that desalination plants would jump at the chance to update their technology. After talking to dozens of potential customers, however, they learned that it would take years--even decades--for those water plants to be able to upgrade their facilities. So the team started contacting people in other sectors. That's when it learned that in the energy industry, companies using fracking techniques to mine natural gas desperately need a way to dispose of brine so it does not enter and contaminate nearby water systems.

A number of investors attended the teams' final presentations. Red Ox team member David Kohn announced triumphantly, "We were blown away by how much these companies will pay for this technology." Indeed, the natural-gas execs were willing to pay 60 times what Red Ox had predicted. The VC guys' jaws dropped.

Of the cohort we taught, 19 of the 21 teams left with commercially viable ideas, which they planned to pitch to private-sector investors. (We should know the results of the second cohort soon.) The program will cost the government about $25 million over two years.

Of course, there will be studies to track the teams' progress. This is, after all, a government-sponsored program. But it's also one that, I'm convinced, works. It doesn't require Uncle Sam to pick winners or losers. And it doesn't put Uncle Sam in the role of venture capitalist. The NSF doesn't invest in these start-ups; it gives these teams what they need to get to the point where they can go out and raise private capital. By the end of the year, 200 teams of scientists will have gone through the program.

We've been sitting on a huge, untapped source of innovative companies and high-paying jobs, something we desperately need. If we do this right, it has the potential to change careers, industries--and even the future of this country.