If the “resounding victory” of artificial intelligence-piloted F-16s over human pilots in a virtual dogfight is any indication, the United States would be at a battlefield disadvantage if China and Russia’s early investments in AI outpace the U.S. commitment.
“We must stay ahead of our near-peer rivals — namely China and Russia,” Defense Secretary Mark Esper declared at the Pentagon’s first Artificial Intelligence Symposium & Exposition Wednesday.
“Those who are first to harness once-in-a-generation technologies often have a decisive advantage on the battlefield for years to come,” he added.
Esper used his short keynote remarks to describe the results of recent Defense Advanced Research Projects Agency tests between experienced American pilots and virtual F-16s piloted by advanced algorithms.
“The AI agent’s resounding victory demonstrated the ability of advanced algorithms to outperform humans in virtual dogfights,” he said. “These simulations will culminate in a real-world competition involving full-scale tactical aircraft in 2024.”
America’s peer adversaries also have aims at fielding AI on the battlefield, Esper added. And some already have.
The defense secretary said Moscow used AI to help decimate the Ukrainian military and that China vowed to become a world leader by 2030.
Despite a $2 billion commitment to AI development two years ago, battlefield tools used by the American Armed Forces are still in the prototype phase.
That’s by design, Nand Mulchandani, the director of the Defense Department’s Joint Artificial Intelligence Center, told the Washington Examiner Thursday.
“I don’t think that there’s a ‘catch up’ there, per se,” he said.
“The U.S. is still ahead compared to China,” he emphasized. “The innovation, both for the DoD and other militaries, is not going to come internally. It’s coming from adopting and using and working with the vendors and the industry out there who are then going to drive this.”
Mulchandani said internet-based advertising is what motivated the private sector to invest heavily in AI technology in recent years.
That foundation has led to AI advances such as self-driving cars, and it will be the foundation for Pentagon efforts to incorporate AI across the services.
“The question becomes, ‘How do we quickly adopt this and harmonically bring this into the DoD the best way?” he said.
Army Col. Brad Boyd, the JAIC’s Chief of Joint Warfighting Operations, told the Washington Examiner it’s not about which country fields fancy weapons first.
“The decisive aspect of AI for warfare is honestly going to be all the unsexy stuff,” he said.
That means building what Mulchandani describes as “infrastructure and platforms,” the technological backbone for widespread Defense Department applications, each with its own “algorithm” specific to the task.
“We’re working the hardest in doing that, that infrastructure, and that’s the key,” said Boyd. “That’s where the fight of 20 years is going to be won. It’s not on an exquisite algorithm right now because that’s going to be obsolete in a year. It’s the infrastructure that allows us to continue to develop innovative algorithms over the next 50 years.”
So far, AI use by the department has been limited, JAIC briefers admitted, some 18 months after the center was stood up.
They include things such as helping Special Operations Command predict engine failures or planning logistics for humanitarian assistance.
“When it comes to battlefield systems, there are very few real AI-based systems out there in the battlefield that are deployed,” said Mulchandani. “We’re still in the very early stages of this.”
Many systems that look autonomous are not in reality, he said.
The MQ-9 Reaper drone, for example, while unmanned, is remotely piloted. The video it gathers is reviewed by humans before a missile is fired at an enemy.
Esper said AI ethics will be fundamental in the department’s ongoing development, and the technology is not meant to replace humans.
“AI’s role in our lethality is to support human decision-makers, not replace them,” Esper said. “We see AI as a tool to free up resources, time, and manpower so our people can focus on higher-priority tasks, and arrive at the decision point, whether in a lab or on the battlefield, faster and more precise than the competition.”
[Read more: ‘I have no desire to wipe out humans’: Robot writes op-ed for the Guardian]
To do that will require a shift in the way Pentagon officials think about the data they collect.
“DoD leaves a lot of data on the floor,” said Boyd.
“Bottom line, we’ve got to get to a point where the Department of Defense is what we would call ‘AI-ready,’” he added. “Being able to collect, store, transfer data, even if you don’t have an application in mind.”
Figuring out how to use different data sets from across the globe, then shortening the time between the event and the automated decision will unlock the value of AI for the military.
“That becomes the core advantage and the thing that wins the game long-term,” explained Mulchandani.
“What you can do then is slide in new AI algorithm systems and iterate on those and swap them in and out based on a modular architecture,” he explained. “But having that platform at the bottom is literally the important point.”
