For a guy who spent the last 20 years either starting new companies or helping large organizations innovate, John Wolpert doesn't talk about innovation very much. Instead, he's more inclined to speak about the merits of talent development, micro-investment, restricting timeframes…and the benefits of drinking too much at tradeshows.

"If a corporate program focuses exclusively on 'innovation', it will be shut down or gamed into irrelevance in two to three years," he says. "If you focus on developing entrepreneurial talent, you can build a program that lasts and delivers more 'innovation' than a pure-play incubator."

Best known for his work at IBM's Extreme Blue incubator for talent, technology and emerging business, Wolpert has spent the past several years running innovation programs in Europe and Australia. Today, he's in California building Team upStart LLC., which helps large companies develop top entrepreneurial talent while executing high-speed, low-cost projects that prove new business opportunities.

After obtaining a degree from UC Berkeley in the late '80s, Wolpert founded a high-tech marketing firm in San Francisco. In the mid-90's, he received his MBA from Georgetown which led to a position at IBM running the company's alphaWorks team in Silicon Valley. It was his first foray into open innovation and the power of collaboration. "alphaWorks was almost a precursor to the notion of Open Source software," says Wolpert. It's success led Wolpert to help build IBM's Extreme Blue program.

Wolpert says: "The reason why Extreme Blue has stayed funded for over a decade is because it isn't an innovation program. It's a talent development program. And that is a perennially central theme in a company like IBM. However, it also produced an astounding number of inventions and was critical in proving many emerging businesses over the past ten years, some of which grew into huge enterprises.

If it were just a talent program, it would have become boring – or a glorified 'ropes course'. If it were just a tech lab, it would have become just another part of R&D. If it were just a business incubator, it would have been shut down long ago. Its power comes from managing that balance, but this is also its challenge."

For all its success, Extreme Blue was not very good at building things that were best run outside IBM, and it suffered from the same problems any large company team has when trying to work with other firms.

In 2003, with the blessing of IBM, he left the company and traveled to Australia to build the Australian InnovationXchange (IXC) intermediary program. The premise was simple: IXC intermediaries would work inside member firms and research organizations under a strict code of ethics—Wolpert calls it a "commercial castration clause"—and systemically search for business partnership opportunities.

"Imagine companies paying hundreds of thousands of dollars a year to have this odd person lurking around looking for stuff and then telling their fellow intermediaries at other companies what they've learned," explains Wolpert. With IBM as his anchor tenant, IXC began building a massive base of thought capital. The company also explored one of the gnarliest innovation dilemmas out there: building trust. "To break down silos, you have to deploy real people who you trust," explains Wolpert. "Intermediaries are go-betweens who are well trained and naturally good at keeping secrets, only making connections with others at the right times and with great discretion. A true intermediary has learned to manage his or her ego and derives no personal satisfaction from being the person who knows something or someone."