The Pentagon's newly established
Algorithmic Warfare Cross Function Team
will deliver an algorithm to the battlefield by the end of the year, its chief has asserted.
Col. Drew Cukor, chief of the Algorithmic Warfare Cross Function Team under the purview of the Office of the Under Secretary of Defense for Intelligence, told an audience at the Defense One Tech Summit in Washington on Thursday that the team’s objective is to field algorithms onto government platforms to lessen the burden of the overworked operator.
Cukor noted that previously in the world of ISR processing, exploitation and dissemination, the focus was on the platform and the sensor. The exploiter was the last consideration for these programs.
“I’m here to announce that that’s changing now,” he said, echoing similar sentiments previously made by Lt. Gen. Jack Shanahan, the head of the team.
Cukor promised there will be an algorithm in a combat zone by the end of this calendar year, but the only way this will be possible is partnering in lockstep with commercial industry.
“I wish we could buy [artificial intelligence] like we buy lettuce at Safeway, where we can walk in and wipe a credit card and walk out with our head of lettuce,” he said. The immediacy of this effort demands that commercial systems ready now be employed, rather than wait five years for the perfect solution to come along.
“You don’t buy AI like you buy ammunition — you don’t just get it,” he continued. There’s a very deliberate workflow process.”
Cukor said the rapid acquisition authorities that the Department of Defense has provided serve as an opportunity for about 36 months to explore the best ways to engage industry in order to advantage the taxpayer and the war fighter, "who frankly wants the best algorithms that exist.”
The goal for this effort, as it currently stands, is not to replace the human operator or allow the machine to make life or death decisions, but create a symbiotic relationship in which the machine aids the human in the decision-making and exploitation process, allowing individual analysts to potentially do twice or even three times as much work as they do now.
This effort does not bring an end to
the department’s policy against machines pulling the trigger
, or allowing a machine to select a target set for a human. “The AI will not be selecting the target anytime soon,” Cukor said. “What the AI will do is complement the human operator.”
There’s not going to be the selection of targets, he said, adding there’s going to be an advisory role, a collaboration role, and detections and alerts — the human is still a fundamental part of this chain.
“When we get to the point where we have decision-making, I believe there will be another healthy debate, but we are not anywhere close to that right now,” Cukor said.