Michael Horowitz of the School of Arts and Sciences weighed in on the possibility of the U.S. adopting an artificial intelligence-based nuclear weapon system. Since these algorithms would depend on the availability of large data sets, which don’t currently exist, Horowitz says “training an algorithm for early warning means that you’re relying entirely on simulated data. I would say, based on the state-of-the-art in the development of algorithms, that generates some risks.”

August 30, 2019 ・ From Bulletin of the Atomic Scientists