U.S. intelligence chiefs plan for a future where every spy uses AI

From Washington Times.com (July 14, 2023):

U.S. spy agencies are preparing for a future where all of its analysts and spies use artificial intelligence to augment their work rather than rely on traditional tradecraft.

The spy agencies’ leadership plans to be “AI-first” and wants everyone from senior leaders on down to leverage artificial intelligence, according to the Office of the Director of National Intelligence’s Rachel Grunspan.

In remarks Friday at the Intelligence and National Security Summit in Maryland, Ms. Grunspan said she foresees a world where spy agencies use AI for hybrid war games, simulations, and strategic “red-teaming” that tests the government’s defenses.

“Anything that is getting AI in the hands of individual officers regardless of their job, regardless of their role, regardless of their background, technical or not, and just maximizing the capacity of the entire workforce,” she said. “That’s where I see us going.”

The spy agencies are not there yet but they are working on it. Lakshmi Raman, the CIA’s AI chief, said her agency is examining the applicability of large language models used by popular tools such as ChatGPT.

“We are in that exploration and experimentation phase,” Ms. Raman said.

The National Security Agency, which is responsible for intercepting electronic and digital foreign intelligence, is also scrutinizing artificial intelligence. NSA’s Jason Wang said his agency is “very active” in studying the opportunities for bringing large language models inside the code-breaking and code-making agency.

Such talk only fan fears that AI will produce a nightmare scenario of robots ruling over humans. This idea has long dominated science fiction, however it now seems realistic.

Billionaire tech entrepreneur Elon Musk is concerned about a future where AI runs the world, rather than the other way around.

Mr. Musk launched an artificial intelligence startup this week called xAI and said on Twitter Spaces that America must worry about catastrophic outcomes.

“It’s actually important for us to worry about a Terminator future in order to avoid a Terminator future,” Mr. Musk said.

OpenAI, the maker of the popular chatbot ChatGPT, is assembling a new team to stop the technology from going rogue and having the potential to extinguish humanity.

The company said last week it did not have a solution for controlling a “potentially superintelligent AI and preventing it from going rogue” so it was assembling a new group of engineers and researchers to solve the problem.

Still, the U.S. appears intent on harnessing superintelligence for its positive potential.

At the conference, Booz Allen Hamilton’s Patrick Biltgen told the AI chiefs of the CIA, NSA, and DNI that he expected the nearest-term application of the new tech would be realized in the form of an AI assistant for all intelligence officers.

Mr. Biltgen, who works on AI at the government contractor with U.S. agencies, said that he envisioned spies using AI similar to how the artificial intelligence character JARVIS assists Tony Stark in the popular “Iron Man” superhero stories.

“I think the analyst-assistant, the JARVIS for your Tony Stark, is going to be the first, most viable thing that’s going to be the viral app for the IC,” Mr. Biltgen said. “It’s like, ‘Jarvis, go get my selectors and put them in a format I need. I’m going to be over here on my other monitor doing this other thing.’” [source]