Read on Twitter
I've been seeing a bunch of "its the end of an era" in response to this article. This is indeed a technical achievement, but its a distraction from where we really need to focus our AI efforts in the DoD (haters stay for the thread).
https://www.defenseone.com/technology/2020/08/ai-just-beat-human-f-16-pilot-dogfight-again/167872/
https://www.defenseone.com/technology/2020/08/ai-just-beat-human-f-16-pilot-dogfight-again/167872/
On the experiment: y'all it was a pilot w/a VR headset & a fake stick. AI beat a human pilot at a video game. It isn't surprising that AI performs well in a simulated environment & that human advantages (the warm fuzzy) are less important. https://www.quantamagazine.org/why-alphazeros-artificial-intelligence-has-trouble-with-the-real-world-20180221/
The transition from this kind of AI to an unmanned platform with integrated sensors, weapons, & combat controls is expensive & vulnerable to both cyber/EM threats. Check out my work w/ @jumacdo on the importance of cost in optimizing unmanned strategies.
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3407957
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3407957
This is my real problem w/this kind of AI theater. The DoD desperately needs AI innovation, but where it can reap the most benefit is not in expensive AI replacements for manned systems but instead AI for health records, personnel management, & logistics. https://www.nature.com/articles/d41586-019-03848-y
But AI for these support tasks is decidedly less sexy than dogfighting & what DoD needs to do to unlock the potential of AI (data management, IT solutions, new policies & practices) is messier, more extensive, & far less interesting than a DARPA spectacle. https://federalnewsnetwork.com/commentary/2020/06/why-congress-dod-should-focus-on-training-data-platforms-to-make-ai-tools-more-valuable/
