Silicon Valley investors to DoD: Dual-use tech is a bad strategy

By: Jill Aitoro

SIMI VALLEY, Calif. — Dual-use technology — that is, tech that can be adapted from the commercial market to serve the needs of the military — is core to the U.S. Department of Defense's innovation strategy.

But those willing to put money toward big ideas argue it's the wrong approach.

“In terms of how to build a startup and how to scale really fast, you can't have two missions,” said Katherine Boyle, an investor with venture capital firm General Catalyst, during a Defense News roundtable in California. “You can't be a 10-person startup saying: ‘OK, we're going to sell to the DoD, but we're also going to sell to these commercial customers, and it's just going to work out magically.'"

For the second year in a row, Defense News hosted the roundtable to dig into Pentagon's efforts to engage with the commercial tech community — this year digging into the challenges and opportunities that come with investment in defense development.

To the Pentagon, dual-use technology offers an attractive means of drawing new players into the military fold, while also leveraging the more rapid development that happens on the commercial side. But the model is evolving, said Mike Madsen, director of strategic engagement with the government's Silicon Valley outreach hub Defense Innovation Unit.

With DoD, “it takes two years to get to a ‘yes,' when a lot of companies need a ‘no' in 30 days because they don't have the capital,” he said. “So we flipped it. Now we start with the DoD problem set and take it out to industry. And we've lowered a lot of the barriers to entry — we negotiate [intellectual property] for each contract, we negotiate auditability, we move quickly. We look to award prototype contracts in 60 to 90 days.”

The approach also attempts to rebalance the gradual shift in research and development investments in the last couple of decades. As noted by Tom Foldesi, DIU's commercial engagement director, one-third of worldwide R&D was tied to the Department of Defense in the 1960s. That percentage has since tanked to 3.7 percent.

A separate business line allows R&D to continue to iterate to the next generation of technology so the DoD can “go back to the cookie jar” and tap into the technology to solve future problems, Foldesi said.

But to Trey Stephens, a partner at venture capital firm Founders Fund and a co-founder and executive chairman of Anduril Industries, the model ensures the large, traditional defense contractors continue to dominate as the small businesses only “dabble in defense.” It also means the DoD won't bear sole responsibility for the economic growth of these small tech startups.

“Where I'm not on board is where a traditional defense company is being asked by the government to integrate dual-use capabilities as a way to prevent that oligopoly from being shaken,” he said. “We have to break this oligopoly. We can only do it if we find companies that are willing to own their responsibility for execution on programs.”

To be clear, Stephens acknowledged cases where commercial technology companies can be primes. Lawsuit aside, he's “on board” with awarding the Joint Enterprise Defense Infrastructure contract to a commercial business — Microsoft — “because the capability is similar enough.” Microsoft was awarded the Pentagon's JEDI cloud contract, but Amazon Web Services has asked a federal court to block the department and the company from beginning work on the project, according to a Jan. 13 court filing.

In terms of new capabilities, Stephens advocates for turning the model on its ear: Enable startups to first development a solution to a problem faced within the DoD, then turn that around and sell it to commercial industries.

“The commercial industry is oftentimes looking to the government for aspirational solutions to some of its hardest problems, whereas the inverse doesn't really work,” he said.

General Catalyst, which counts The Honest Company, Snapchat and Airbnb among its portfolio of companies, has invested in two pure-play defense companies: Anduril, and Palo Alto machine-learning company Vannevar Labs. The latter is developing a product that would bring natural language-processing technologies to support counterterrorism missions.

“We actually think this is a better model,” Boyle said. “If you're scaling rapidly, you have to be very focused on your customer set. And if you're going to have to sacrifice a customer, even if you're a multibillion-dollar company, you're going to sacrifice the one who's moving the slowest. And that's usually the government.”

https://www.c4isrnet.com/smr/cultural-clash/2020/01/30/silicon-valley-investors-to-dod-dual-use-tech-is-a-bad-strategy/