Situational Awareness of Artificial Intelligence – a Military Perspective

Today is the Fourth of July and the United States celebrates itself as the beacon of freedom and democracy in the world. Leopold Aschenbrenner originally hails from Germany and now lives in San Francisco, California. His “aspiration is to secure the blessings of liberty for our posterity.” To that end he wrote a series of papers on Situational Awareness: The Decade Ahead. Artificial Intelligence will explode in magnitude and reach over the next 3 years. Like Aschenbrenner, I originally hail from Germany, and I also chose to make the United States my home and I want to preserve it as beacon of freedom and democracy in the world. On this holiday, I am reviewing the military implications that AI will bring to bear according to Aschenbrenner.

Situational Awareness
Situational Awareness is a military term that roughly encompasses a commander’s knowledge of the operational environment. Aschenbrenner delivers a 30,000 ft overview that accomplishes that for any leader, military or otherwise.

Military Decisions
More precisely Aschenbrenner helps his readers understand the operational environment as it is shaped by AI. In the military decision-making process (MDMP) staff constantly check the facts and assumptions that go into their mission analysis. A disturbing and enlightening take-away from Aschenbrenner’s writing is that most of AI capabilities that I have regarded as assumptions are actually facts already. Aschenbrenner reports confidently in the present tense, “it’s no longer about estimates of human brain size and hypotheticals and theoretical extrapolations and all that—I can basically tell you the cluster AGI will be trained on and when it will be built, the rough combination of algorithms we’ll use, the unsolved problems and the path to solving them, the list of people that will matter. I can see it.” (page 159)

Challenges
Aschenbrenner is neither a doomsday prophet nor a naïve cheerleader. He is brutally honest in listing the risks that AI will pose in the very near future:

• Superintelligence will “provide a decisive and overwhelming military advantage” (page 70) to the actor who develops it first.
• Superintelligence will “be able to overthrow the US government.” (page 70)
• “More generally, everything will just start happening incredibly fast. And the world will start going insane. Suppose we had gone through the geopolitical fever-pitches and man-made perils of the 20th century in mere years; that is the sort of situation we should expect post-superintelligence. By the end of it, superintelligent AI systems will be running our military and economy. During all of this insanity, we’d have extremely scarce time to make the right decisions.“ (page 72)
• “It’s hard to overstate how bad algorithmic secrets security is right now. Between the labs, there are thousands of people with access to the most important secrets; there is basically no background-checking, silo’ing, controls, basic infosec, etc. Things are stored on easily hackable SaaS services. People gabber at parties in SF. Anyone, with all the secrets in their head, could be offered $100M and recruited to a Chinese lab at any point.” (page 96)
• “Eventually, the decision to greenlight the next generation of superintelligence will need to be taken as seriously as the decision to launch a military operation.” (page 122)

Opportunities
On the plus side, if US policymakers get it right on time there are great benefits:

• “It seems likely the advantage conferred by superintelligence would be decisive enough even to preemptively take out an adversary’s nuclear deterrent. Improved sensor networks and analysis could locate even the quietest current nuclear submarines (similarly for mobile missile launchers). Millions or billions of mouse-sized autonomous drones, with advances in stealth, could infiltrate behind enemy lines and then surreptitiously locate, sabotage, and decapitate the adversary’s nuclear forces.” (129-130)
• The MDMP will be enhanced by adding AGI models “as smart as PhDs or experts that can work beside us as coworkers.” (page 9) They may not look like Star Trek’s Data, but they will function in a similar manner.

Happy Birthday, America!
Aschenbrenner released this series shortly before the Fourth of July. And I wholeheartedly concur with his parting admonition, “the free world must prevail over the authoritarian powers in this race. We owe our peace and freedom to American economic and military preeminence. Perhaps even empowered with superintelligence, the CCP will behave responsibly on the international stage, leaving each to their own. But the history of dictators of their ilk is not pretty. If America and her allies fail to win this race, we risk it all.” (page 136)