Amid larger and increasingly more sophisticated Chinese military exercises in the Western Pacific and especially near Taiwan, the U.S. needs better data analytic capabilities to discern intent and provide warnings, the Biden administration’s nominee to command U.S. forces in the Indo-Pacific region said this week.

“First, mass data analytics” that “go deeper to be able to determine strategic, operational, and tactical warning of the force,” Adm. Samuel Paparo, commander of the U.S. Pacific Fleet, said Tuesday afternoon in Silicon Vally during a fireside chat hosted by Doug Beck, director of the Pentagon’s Defense Innovation Unit (DIU). Paparo is awaiting Senate confirmation as commander of U.S. Indo-Pacific Command.

China’s military activity in the past few years has demonstrated “step level changes” in terms of force levels and joint operations that soon become able to “execute a profound military operation…operating under a fig leaf of exercise,” Paparo said. The size and complexity this activity will make it difficult to be forewarned of an actual threat, he said.

Paparo’s warning of potential Chinese military aggression sprung from the guise of a large-scale exercise is given life in White Sun War, a new novel by strategist and retired Australian Army Maj. Gen. Mick Ryan, who paints a scenario of a surprise attack against Taiwan in May 2028 just when it appears a major annual People’s Liberation Army exercise in the Taiwan Strait is concluding.

This challenge requires resources to ultimately get “indications and warnings” from the data, Paparo said. The data, and computing power to make sense of the information, “gives us some probability or some forecast of activity, and then preferably with recursive loops within it, that allow our own algorithms to make those determinations [and] get better and smarter based on behavior that comes,” he said.

Paparo told Beck that a second gap to be filled is with disruptive technologies, which he described as autonomous unmanned systems that include artificial intelligence and machine learning to enable “swarming.” Unmanned systems that have these capabilities can deny an enemy that is trying to take space and do it at a lower cost, first in terms of human life, followed by dollars, and then the opportunity cost, he said.

The proliferated use of unmanned systems equipped with sensors “gives us a stare instead of blink,” Paparo said. AI and machine learning and swarming further the effectiveness of unmanned systems instead of relying solely on a human operator, he said.

Later during the DIU Summit, Bob Stephenson, executive director for maritime information warfare for the Pacific Fleet, said unmanned systems are needed for one-way strike missions, to serve as decoys to take fires, perform resupply operations, and other missions and be resilient against enemy jamming or cyber attacks.

Paparo wants these highly capable unmanned systems in the hands of warfighters to experiment with and to “unlock their creativity.”

Paparo offered a “dictum” on the role of humans and machines as autonomous technology continues to improve.

“All anybody wants to do is they want to solve problems with manpower,” he said. “And we should have a dictum that we should never have a human being doing on graph paper what a machine can do better and faster for us. And we should never abdicate human responsibility to a machine. And these two mutually supportive ideas should inform.”