GenAI could be key to determining military advantage

Eric Schmidt
U.S. Under Secretary of Defense for Acquisition and Sustainment Ellen M. Lord is joined by Defense Innovation Board members s Dr. Eric Schmidt, Dr. Michael McQuade, and Richard Murray to speaks to reporters on the Defense Innovation Board's final report of the year-long Software Acquisition and Practices analysis at the Pentagon, Washington, D.C., May 3, 2019. (DoD photo by Army Sgt. Amber I. Smith)

The Pentagon has been testing different large language models (LLMs) for most of this year with a view to developing digital platforms that leverage a variety of defence data sources. Meanwhile, Deputy Secretary of Defense Dr. Kathleen Hicks oversaw the setting up of Task Force Lima, launched in August, charged with assessing, synchronising, and employing generative AI capabilities across the DoD.

The DoD is moving fast to embrace generative AI, but there are some that believe it should move even faster. In September, the Special Competitive Studies Project (SCSP) urged the Pentagon to establish a Defense Experimentation Unit (DEU) to “provide for much-needed operational experimentation and iteration with AI models across missions, in order to enable their broader and faster deployment and mainstreaming across the Joint Force”. The SCSP recommends that the DEU is stood up by January.

Although a non-profit, chaired by Eric Schmidt, previously chairman of the National Security Commission on AI and former Google CEO, the SCSP’s recommendations carry some weight. The think tank argues that GenAI will accelerate the key trends that will help to determine U.S. military advantage, in particular vis-a-vis China, and so needs to be elevated in priority itself.

According to SCSP, generative AI will accelerate “the imperative of human-machine collaboration and teaming”, the criticality of software advantage, and the necessity of “empowering warfighters at the edge so that they can conduct distributed, network-based operations”.

Empowering warfighters at the edge is without doubt one of the hardest nuts to crack.

U.S. forces have recently made efforts to accelerate their integration of data platforms and use of AI to process data from the ever increasing number of data sources in theatres of operation, so informing decision making. Advances have been made too, in developing networks and devices for those at the frontline to receive data, insights and software apps.

However, huge volumes of data are now collected from space, air, land and sea during conflicts, not to mention the data and communications sent from military personnel themselves. Therefore, analysing relevant data in a timely fashion and using it to empower, rather than burden or confuse warfighters is a significant challenge.

What if GenAI could help solve that?

by Carrington Malin