The US National Security Commission on Artificial Intelligence (NSCAI) has released its final report on the current state of AI development in the US and the threats posed by China’s rapidly developing AI capabilities.
As the 750-page report notes, “China possesses the might, talent, and ambition to surpass the United States as the world’s leader in AI in the next decade if current trends do not change.”
The commission is also concerned that AI is deepening the threat posed by cyberattacks and disinformation campaigns from Russia, China, and others.
“The limited uses of AI-enabled attacks to date represent the tip of the iceberg,” the report warns. “Meanwhile, global crises exemplified by the COVID-19 pandemic and climate change highlight the need to expand our conception of national security and find innovative AI-enabled solutions.”
Former Google CEO Eric Schmidt and chair of the NSCAI outlined four key proposals the US needs to embark on, including a call for the US to spend $40bn to expand and democratize federal AI research and development (R&D).
The $40 billion spending recommended for expanding and democratizing federal AI R&D is a “modest down payment on future breakthroughs”, the report states.
The 15 member commission is proposing to set up a technology competitiveness council at the White House and wants the Department of Defense to be prepared for AI competition.
Schmidt said the US had a “huge talent deficit” and that the country needs to develop new talent and expand existing programs in government. The US also needs the world’s best talent to stay in the US to cultivate homegrown talent, he said.
The third key point is developing hardware and addressing the US’s lack of semiconductor manufacturing capabilities.
“Its really important that our hardware stay ahead, be cutting edge and we are very close to losing the cutting edge in microelectronics,” said Schmidt. “We need to revitalize domestic semiconductor manufacturing and ensure that we’re two generations ahead of China.”
The fourth proposal is around AI research and the need for a massive boost in investment.
“AI research is going to be incredibly expensive so we need the government to help set up the conditions for accessible domestic AI innovation,” said Schmidt.
Schmidt said the US needs a national AI infrastructure. “We need more money. I’m sorry to say, but we do need more money, particularly in AI R&D, so that by 2026 we get to $32 billion per year,” he said.
The report also recommends the Pentagon establish the foundations for widespread integration of AI by 2025. It requires building common digital infrastructure, developing a digitally-literate workforce, and instituting more agile acquisition, budget, and oversight processes.
The Pentagon should divest from military systems that are ill-equipped for AI-enabled warfare and instead invest in next-generation capabilities, it recommends.
