Edward Geist

Exciting news! My book _Deterrence under Uncertainty: Artificial Intelligence and Nuclear Warfare_ is now out in paperback!

A new @RANDCorporation paper by @sovietologist argues that even the most advanced AGI will be constrained by thermodynamics, information theory, and computational complexity. rand.org/pubs/working_p…

🔥Edward Geist from the RAND Corporation discusses the nuclear strategies of 🇷🇺Russia, the 🇺🇸United States, and 🇨🇳China, as well as the role of nuclear power in the era of artificial intelligence. The interview also covers the nuclear policy that 🇵🇱Poland and 🇺🇦Ukraine should adopt. Wersja z polski lektorem już wkrótce. 🔥youtu.be/ctaToIs5IsQ @sovietologist @RANDCorporation

@RogerGrosse This 2011 piece by Pat Langley offers some useful historical perspective on how ML came to focus increasingly on benchmarks starting in the late 1980s: isle.org/~langley/paper…

Is there a good history of how academic AI research came to be so focused on hill-climbing on benchmarks? Any classic essays championing this as a good way to make progress?

This spring, the U.S. and China are expected to start talks on AI safety. Expectations are low, says RAND's @sovietologist. But early, "unproductive" talks aren't a total loss—as illustrated by nuclear arms negotiations during the Cold War. bit.ly/4cIVcIN

"If I had to pick one single problem that causes me the most anxiety, it's North Korea," RAND's Edward Geist tells @1A. North Korea may build up its arsenal and become "a kind of super rogue" that doesn't play by the same rules as major nuclear powers. the1a.org/segments/the-r…

What does today's nuclear landscape look like? How could AI affect nuclear warfare? And what role do scholars play in understanding (and mitigating) ever-changing threats? RAND's @sovietologist discusses on @1A: the1a.org/segments/the-r… (starts at 29:55)

💥📚 I read this #book by Edward Geist a few days ago and it is #excellent. Thank you very much for writing it @sovietologist. 🔴global.oup.com/academic/produ…

Uh oh. Edward Geist, “Deterrence Under Uncertainty, Oxford, 2023, p 6

@mjgault interviewed me about my new book for the Vice's Cyber podcast vice.com/en/article/epv…

We're looking for the next generation of thought leaders on nuclear security. Applications for the Stanton Nuclear Security Fellows Program are now open. Learn more and apply by January 19: bit.ly/3Nie2LX

Sci-fi has convinced us that a Terminator-like Skynet is closer on the horizon than it actually is. It's a great story premise, says analyst Edward Geist. “But it’s also the case that there’s effectively no one who thinks of this as a great idea.” #AI codastory.com/authoritarian-…

@RANDCorporation @sovietologist @NPR Great book. Changed my perspective on the most effective use of AI for nuclear deterrence. Highly recommend.

@sovietologist I have my copy!

@nprscottsimon interviewed me about my new book Deterrence under Uncertainty: Artificial Intelligence and Nuclear Warfare npr.org/2023/12/09/121…

I'd like to thank everyone who has bought the book already, as it is currently sold out in the U.S.! Oxford University Press assures me that more copies are being shipped over from the UK global.oup.com/academic/produ…

Now out in the U.S.!

I began working on this project in 2017 and I would like to express my gratitude to everyone who helped bring it to fruition, especially the Smith Richardson Foundation for supporting it with a Strategy and Policy Fellowship.

My new book Deterrence under Uncertainty: Artificial Intelligence and Nuclear Warfare is out in the U.S. today! global.oup.com/academic/produ…