Stanford remains Silicon Valley’s singular institution—the root of its vibrancy and hegemony. And because of Stanford Research Park, whose original tenants included Lockheed Martin and other defense contractors, it’s also a physical reminder of the huge role defense played in the early life of the Valley. Even though the industry around it has moved on to smartphones and social networks, a popular course at the university has made Stanford’s defense connection current again.
In Hacking for Defense, students spend the semester in teams and are challenged to build tech products that address actual problems submitted by defense and intelligence agency sponsors. The class was started in 2016 by Stanford historian and entrepreneur Steve Blank, ex-Army Colonel and current BMNT CEO Pete Newell, and Joe Felter, a former U.S. deputy assistant secretary of defense and current Hoover Institution fellow. The class turned into a federally sponsored program the following year, funded by the National Defense Authorization Act. You can now find Hacking for Defense classes following the Stanford model at 55 U.S. universities. As of last spring, 2000 students had taken the H4D course.
Blank says that Americorps and the Peace Corps shouldn’t be the only options for students interested in public service; they should be able to contribute their time and talent to solving difficult defense- and intelligence-related challenges, too.
Now, thanks to its Hacking for Defense class, Stanford is minting future startup founders and executives that could eventually enlarge the tech industry’s presence in the country’s defense industrial base. This is especially relevant now that the nature of warfare is changing. No longer are wars fought and won merely on the strength of big weapons and brute force: The deciding factors in modern war are more diverse, spanning autonomous weapons, information warfare, bioweapons, and war in space.
During Obama’s second term, people in the Pentagon realized that Uncle Sam would need the help of the country’s best tech talent to defend against things like drone swarms, hypersonic missiles, cyberattacks, and disinformation campaigns. The Pentagon began sending people to Silicon Valley to make connections. In 2015 it set up the Defense Innovation Unit in Mountain View to find cutting edge tech that could be relevant to defense needs. At Stanford, it felt like history repeating itself. At the beginning of the Cold War, Stanford was considered the premiere research institution for developing defense technology. In the 1950s, Stanford science and engineering grads would go on to launch the Valley’s first silicon chip makers.
That legacy continues through Hacking for Defense. Sitting in on the Hacking for Defense class in spring 2021, which was still remote via Zoom, I was taken aback by the seriousness of both the students and their projects. I watched Team AngelComms present their progress creating a small communications device that would help in the rescue of downed pilots. Team Radicalization Inoculation described the lessons it learned about the market-fit of its product proposal—software that lets the service branches track people within their ranks with suspected ties to domestic terror groups. An all-female team of students, Flexible Fingerprints, reported on its prototype software to accelerate cybersecurity software integration into DoD servers. That team is now pursuing VC funding.
H4D instructor Jeff Decker tells me 20 teams typically apply for admission to the H4D course, from which the teaching team selects eight to ten. A third of the students come to the course come from the engineering school, a third from the business graduate school, and a third from studying law, policy, or international relations, Decker says.
Each week, the teams call on at least 10 end users within the sponsor agency and gradually evolve their product plan to better fit the agency’s need. Many of the presenters I saw weren’t even 20 years old, yet they spoke like veteran product managers. Blank’s main feedback to them was to slow down: The students were so immersed in their project and market research that many of them raced through their presentations.
Blank and his fellow instructors, which include both other professors and defense-industry professionals, stressed that while product planning is important, students also must zero in on the right kind of DoD funding type and identify the champions and decision-makers within the agency that can push the contract through.
Nationwide, Hacking for Defense programs have birthed at least 14 startup companies. These include the satellite imagery company Capella Space, flexible lithium-ion batteries developer Anthro Energy, and Lumineye, which develops radar systems that may help soldiers see through walls.
Team SilkNet built its product around this challenge from the Army Research Lab: “Platoon Radio Control operators need effective human-machine teaming for enhanced situational awareness.” SilkNet developed a small “leave-behind” sensor set contained in a “sticker” that can detect people, vehicles, and electromagnetic signals—all without relying on a network connection.
The week before the class, SilkNet had an important conversation with a procurement officer at the Special Operations Command, or SOCOM. The officer said SOCOM had been looking for the kind of product the team had developed, that it “could and would” buy it. But there was one caveat. The product had to be fully developed and “ready to hand over.”
SilkNet wasn’t there yet. A lot of prototyping and sensor integration needed to be done, probably by engineers that the team would need to hire. To make the product ready for a potential buyer like SOCOM, SilkNet is applying to get a Small Business Administration loan to fund its development for defense. This isn’t unusual—the class teachers encourage the teams to continue working on their projects after the class is over, and perhaps found their own startups.
The students in the Hacking for Defense class seemed free from the moral quandaries of tech employees encountering defense work for the first time at companies like Google. They come with open eyes to solve issues which could have geopolitical significance.
“If you’re an engineer or scientist, and you want to work on the toughest problems in the world, trust me, [you’re] not working for Google,” Blank says, using the company as shorthand for the mainstream tech industry. “Some of the toughest problems are solving real AI, real autonomy, real whatever—problems to keep the country safe and secure.”