WASHINGTON — ChatGPT, a bot launched by OpenAI in November that can generate human-like conversations and content, including surreal art and computer code, has caught the attention of U.S. defense officials.
According to Chief Technology Officer Stephen Wallace, the technology that underpins the viral bot, known as generative artificial intelligence, was recently added to the Defense Information Systems Agency’s watch list.
“We’ve heard a lot about AI over the years, and it’s already working in a lot of places,” he said. 25th event hosted by a chapter of the Communications and Electronics Interest Group AFCEA at the Army-Navy Country Club in Arlington, VA. “But this ability to generate content is a very interesting one.”
The watchlist is updated regularly and has featured topics in the past that have since become pillars of defensive connectivity and security, such as 5G, zero-trust digital defense, quantum-resistant cryptography, edge computing and telepresence.
“We set out to study: how [generative AI] It actually changes DISA’s mission in the sector and what we provide to the sector,” Wallace said.
ChatGPT, praised by some for its potential to improve employee productivity but scorned by others for its bias and ethical concerns, surpassed 1 million registered users within a week of its launch. Bill Drexel, an associate fellow for technology and national security at the Center for a New American Security, said the easily accessible platform demonstrated the power and flaws of artificial intelligence to a “large population.”
“While it’s clearly not a military system per se, I think the increased exposure to these occasional, often enterprise-driven enterprises does raise awareness of what these tools can do right and what can go wrong,” He told Defense News on Jan. 1 in a separate livestreamed event. 26.
Exactly how the Pentagon will apply generative AI is unclear, and Wallace did not provide specifics. OpenAI CEO and ChatGPT creator Sam Altman met with lawmakers this week to try to demystify the tool, Semafor reported.
The U.S. military is increasingly investing in artificial intelligence and related technologies as a means to improve applications such as battlefield analysis and predictive maintenance needs.
Pentagon public spending on AI (including autonomy) has ballooned from just over $600 million in 2016 to $2.5 billion in 2021. More than 685 AI projects are underway, including several related to major weapons systems, according to data from the beginning of 2021, according to the Government Accountability Office.
In November, Air Force Chief Information Officer Lauren Knausenberger said the Air Force must “automate more” to remain dominant in a world of advanced computing and lightning-fast decision-making.
Colin Demarest is a reporter for C4ISRNET covering military networks, cyber and IT. Colin previously covered the Department of Energy and its National Nuclear Security Administration—namely, Cold War cleanup and nuclear weapons development—for a South Carolina daily. Colin is also an award-winning photographer.