The Trump administration expects to make a “significant increase” in research spending for artificial intelligence and quantum information science next fiscal year, and has proposed doubling the amount of non-defense R&D spending in those areas by FY 2022.
Increased funding for federal AI and quantum research, however, would come at the sacrifice of spending on other areas of research. The Trump budget request proposes cutting total federal R&D spending by nearly $13.8 billion, a 9% cut from estimated FY 2020 spending levels.
“I think the key priority of this administration is prioritizing research and placing our money on things that are extremely important,” said Kelvin Droegemeier, the director of the White House Office of Science and Technology Policy, in a call with reporters on Monday.
As for R&D in areas deemed a “lower priority” for the federal agencies, Droegemeier suggested private industry and academia could step up their investments in those sectors.
In reality, however, Congress has much more final say in funding these and other federal programs and often sets those spending levels without much regard to any president’s largely symbolic budget requests.
As for what federal agencies actually spent on AI research last year, the White House expects to release those numbers later this year.
U.S. Chief Technology Officer Michael Kratsios said the White House’s Networking and Information Technology R&D (NITRD) program would release a budget supplement this summer detailing how much agencies spent on AI research in FY 2019.
The document, he added, would also include a “definitive cross-cut” of agency-by-agency spending on AI research for FY 2021.
The FY 21 budget supplement on AI R&D won’t be available until this summer, Kratsios said, because the budget analysis is “extraordinarily complicated and taxing on the agencies to pull together.”
To put agencies on track to double their AI R&D spending by FY 2022, the Trump administration has outlined the following agency-specific proposals:
The budget would increase AI R&D spending at the National Science Foundation and “interdisciplinary research institutes” to more than $850 million, nearly a 70% increase from the FY 2020 budget request.
The Energy Department’s Office of Science would invest $125 million in AI research, a $54 million increase over the FY 2020 budget request.
The Agriculture Department would provide $100 million in competitive grants through its Agriculture and Food Research Initiative, which looks to bring more emerging technology such as AI into agricultural systems.
The National Institutes of Health would invest $50 million in new research on chronic diseases using AI and other technologies.
Rep. Will Hurd (R-Texas), former chair of the IT subcommittee of the House Oversight and Reform Committee, said the budget request reflected the administration’s priorities on emerging technology.
“America must lead in artificial intelligence, quantum computing and other technologies that will define the 21st century,” Hurd said in a statement Tuesday. “I applaud the administration for recognizing the importance of AI and quantum research and will continue to work with my colleagues to prepare our nation for these challenges and ensure America remains the most important economy in the world.”
Under the president’s budget proposal, DoD’s AI budget, located within the Office of the Undersecretary of Defense for Research and Engineering, would increase from $780 million this year to $841 million in FY 2021.
“The U. S. cannot afford to lose our global leadership position in emerging and disruptive technologies, such as AI, which are instrumental in maintaining our strategic competitive advantage,” JAIC Director Lt. Gen. John Shanahan told reporters Monday.
The Defense Advanced Research Projects Agency’s AI research, he added, “represents a significant portion” of next year’s budget request.
DARPA projects an investment of $459 million in AI R&D next year, a $50 increase from this year.
“DARPA is aggressively applying current AI technologies to the hardest DoD problems, establishing science and tools to assure robust, trusted performance of current AI and leading the creation of the next wave of AI technologies, which are fundamentally different,” Shanahan said.
Meanwhile, the JAIC, which serves as the “focal point” for DoD’s effort to accelerate and scale up the use of AI capabilities in the field, would see its AI spending increase from $242 in FY 2020 to $290 million in FY 2021.
The JAIC’s portfolio of AI use cases for the year ahead touches on everything from warfighter health, to humanitarian assistance, to disaster response and joint warfighting.
Its most notable project, however, remains Project Maven, an initiative that brought together some of the country’s top tech companies to develop machine learning tools that could sort through drone footage to support counterterrorism operations.
Shanahan said Project Maven’s capabilities have been fielded to all of DoD’s military services.
“Maven will continue to expand its lines of effort into other intelligence-rated projects. We have the solemn responsibility to place the world’s best technology into the warfighters,” he said. “AI funding in this budget will accelerate our efforts to partner with industry, academia and our allies and partners to ensure we do exactly that.”
In the final draft of its recommendations, the board of tech industry experts strongly urged DoD to have the ability to pull the plug on AI systems, just in case those algorithms begin to make decisions that are beyond their pay grade.
While a final decision has not yet been made, Shanahan said DoD expects to “make recommendations that would largely support taking the DIB principles and turning those and promulgating them as department AI principles.”