A long-running Senate task force has released policy recommendations for federal funding of artificial intelligence: $32 billion per year, covering everything from infrastructure to grand challenges to national security risk assessments.
This “road map” is not a bill or detailed policy proposal, but it provides a sense of the scale at which lawmakers and “stakeholders” are considering when approaching real problems – albeit in an election year. The probability is so small that it almost disappears.
In a final report released by the office of Sen. Chuck Schumer (D-N.Y.), the bipartisan task force identified the most important areas for investment to maintain U.S. competitiveness against overseas competitors.
Here are some important items on the roadmap:
- “Across-government AI research and development efforts, including related infrastructure,” means engaging the Department of Energy, the National Science Foundation, the National Institute of Standards and Technology, NASA, the Department of Commerce and six other agencies and departments with AI-friendly ways to format and share data. In a way, this relatively simple-sounding task is the most arduous of them all and may take years to complete.
- Fund U.S. AI hardware and software efforts at the semiconductor and architecture levels through the CHIPS Act and other means.
- Further fund and expand national AI research resources that are still in their infancy.
- The “Artificial Intelligence Grand Challenge” uses competition to stimulate innovation. “The application of artificial intelligence will fundamentally change the process of science, engineering or medicine, as well as the basic topics of safe and efficient software and hardware design.”
- “Support AI readiness and cybersecurity” in elections, specifically “reducing objectively false content generated by AI while still protecting First Amendment rights.” It might be harder than it sounds!
- “Modernize the federal government and improve the delivery of government services by updating IT infrastructure to leverage modern data science and artificial intelligence technologies and deploying new technologies to identify inefficiencies in U.S. regulations, federal rules, and procurement programs” “. I understand what they are saying here, but this is a big challenge for artificial intelligence programs.
- Lots of vague but large-scale defense-related things like “assessment and mitigation of chemical, biological, radiological, and nuclear (CBRN) AI-enhanced threats by the Department of Defense, Department of Homeland Security (DHS), Department of Energy, and other related agencies.”
- Investigating ‘regulatory gaps’ in finance and housing where AI-driven processes could be used to further marginalize vulnerable groups.
- “Review whether other potential uses of artificial intelligence should be severely restricted or prohibited.” Following a section on potentially harmful content such as artificial intelligence-driven social scoring.
- Legislation banning artificial intelligence-generated child sexual abuse material and other non-consensual images and media.
- Ensure that NIH, HHS, and FDA have the tools they need to evaluate artificial intelligence tools in health care and medical applications.
- “Establishing a coherent approach to meeting public transparency requirements for artificial intelligence systems,” whether private or public.
- Improve the general availability of “content source information” (i.e., training data). What are the models made of? Do you use this model to train it further? etc. Artificial intelligence manufacturers will fight tooth and nail until they can adequately clean up the illegal material used to create today’s artificial intelligence.
- Look at the risks and benefits of using private AI versus open source AI (if the latter exists in a scalable form).
You can read the full report here; the content above (a longer list than I expected to write) is drawn from more bullet points. No budget figure was suggested.
Given that the next six months will be spent largely on election-related tedious work, this document is more about rooting in a lot of general ideas than stimulating actual legislation. Much of what is proposed will require months or even years of study and iteration before a law or rule is enacted.
The AI industry is growing faster than other tech industries, which means it’s orders of magnitude faster than the federal government. While the priorities listed above are mostly cautious, one wonders how many of them will still be relevant when Congress or the White House actually takes action.
