Defense industry must take lead on AI as tech firms waver

In September 2018 the Pentagon announced it would invest up to $2 billion over five years in development of artificial intelligence (AI).

The investments will be managed through the Defense Advanced Research Projects Agency (DARPA), an agency that has a respectable track record in artificial intelligence research and development. Most celebrated the news as a sign that the United States was ready to compete in the growing AI arms race.

{mosads}Of course, the Chinese have a national strategic plan that will commit as much as $7 billion by 2030. They’re spending $2 billion alone on a dedicated office park to centralize all their efforts. The audacious ambition of the Chinese plan has triggered a bit of panic in the defense industry.

The U.S. has no such plan, though we’re still widely perceived to be the undisputed leader in artificial intelligence. Not to mention, the U.S. has the most advanced academic research institutions in the world, and a private tech sector that continues to be the standard bearer for technology innovation. 

The U.S. system for research and development (R&D) relies heavily on universities and private corporations, a strategy that consistently puts us at the top of the list of the biggest spenders in R&D.

The U.S. spent $553 billion in 2018, which amounts to 2.84 percent of gross domestic product. China comes in second at $475 billion, representing only 1.97 percent.

It would seem the fearful reaction to the Chinese plan is, at least, misplaced. The U.S. will continue to lead in AI and overall R&D investment for the foreseeable future. The real question is: Which players in the U.S. system should lead the charge against efforts like the centralized and nationalized Chinese strategic plan? 

To date, Silicon Valley has been a leader in AI. However, when Google backed out of Project Maven, it sent a signal that some of the largest players in Silicon Valley were not interested in leading from the front with their defense partnerships.

If the largest, most powerful companies in Silicon Valley aren’t willing to lead then it’s time for the defense industry set the agenda. 

In June 2018, then-Deputy Secretary of Defense Patrick Shanahan issued a memo establishing the Joint Artificial Intelligence Center (JAIC).

The JAIC will be guided by the goal of “accelerating the delivery of AI-enabled capabilities, scaling the department-wide impact of AI and synchronizing DoD AI activities to expand joint force advantages.” Notably, the deputy secretary reaffirmed the department’s commitment to “military ethics and AI safety” in all of its efforts.

While the private sector is wise to consider the ethical implications of the partnerships they strike, it’s critical for technology companies to establish clearer expectations for engagement in defense projects.

As a country, we depend on the private sector to develop and implement technologies in service of our national defense (see here, here, and here), and the Pentagon has consistently upheld well-established ethical standards for the use of emerging technologies, such as the computer vision employed in Project Maven.

By definition our defense agencies operate with more oversight, checks and balances than any company in the private sector. Institutionalized oversight provides consistent, predictable terms of engagement for technology companies, and buyers in the defense industry should hold vendors to the same standard.

Those watching the global AI arms race would be wise to worry less about the investments of the U.S. government and instead pay close attention to the American private sector, where the bulk of our intellectual property and implementation capacity resides.

If the Silicon Valley elite continue to get cold feet around defense contracts, it could have systemic effects not only on private-sector innovation, but on the competitiveness of our national defense. 

As we embark on a new age of military technology, we’re faced with the daunting task of rewriting our entire approach with machine intelligence playing a pervasive role in areas that range from cybersecurity to weapons, operations and intelligence, surveillance and reconnaissance.

Existing private-sector technologies, including computer vision, machine intelligence and even augmented reality, are mature and ready for implementation. The question is no longer technological readiness, but operational readiness.

Can the private sector deploy these technologies successfully in a defense setting? Can Silicon Valley juggernauts mature their organizations to better serve these customers within the scope of their corporate missions and values? 

{mossecondads}Fortunately, the defense industry and the thousands of contractors who have served our national defense agencies for decades have ample institutional knowledge to deliver our most advanced AI, machine intelligence and computer vision technologies to the front lines of our national defense.

It’s time for the private tech sector and the defense industry to forge partnerships that not only strengthen our world-leading R&D efforts but also accelerate the adoption of these technologies in our defense agencies.

The appropriate response to nationalized AI strategy is not to abandon the distributed approach that put us ahead of all other nations. It’s certainly not a time to argue about the future of AI in the military.

Instead, we need the private sector to step up with urgency and to work with the defense industry to deliver solutions that ensure the world’s most important military is able to protect the world’s most important engine of progress: American innovation.

Ben Lamm is a serial software and tech entrepreneur and investor. He is the founder of Hypergiant, Conversable, Chaotic Moon Studios and Team Chaos. Based in Austin, Texas, he was named one of Adweek’s Top 10 Rising Tech Stars of 2017.