We’ve seen the rise of unmanned vehicles that can accomplish a mission just as well as a manned vehicle can. It is keeping our service men and women out of harm’s way while still kicking butt on the battlefield. Seems like a win/win for us and a recent $2 billion investment by the government into research and development of AI for battlefield use is intensifying off a global AI arms race.

Military value of AI

AI has great military value. We all know that computers can accomplish many tasks quicker than humans can. With AI bombing of a target can be done much more precisely, artillery strikes can be much more limited and unmanned drones are gathering reconnaissance information without putting a human at risk. There is great potential with AI and we are not alone in recognizing it as both Russia and China are investing heavily in it as well.

Several projects are already underway through DARPA (Defense Advanced Research Projects Agency). Contractor Booz Allen Hamilton is working on undescribed AI projects (we’re not on a need to know basis) and Google is working on Project Maven which will improve a computer’s ability to pick out objects in an image, which has been controversial as several employees publicly voiced their displeasure with the project.. It is believed that one goal set forth for this entire investment is to develop algorithms for technology that can be installed on weapon systems that can make decisions on their own and kill.

The future of war

AI systems are already capable of selecting targets and firing but many commanders have been reluctant to give the machine full control. It is an unproven system and the humans lack confidence in the machine’s ability to reason. The goal of the projects is to get AI to the point where it can explain its reasoning while on a battlefield. Many are worried about the adaptability of the algorithms and what would happen if they malfunction or are not programmed correctly. As Michael Horowitz, who worked on AI for the Pentagon says, “It’s one thing if what you are talking about is a Google search, but it’s another thing if what you are talking about is a weapons system.”

AI on par with the Manhattan Project

The endgame is essentially to get AI to have common sense. The Pentagon has been working on that for well over a decade and the $2 billion investment in this go around is considered to be a drop in the bucket compared to what will be needed to even develop a working AI system on par with a young child. One former DARPA manager said that what will be needed is a Manhattan Project style budget to even begin to make progress. Considering that the Manhattan Project cost about $2 billion in 1940s money ( $22 billion today) this will not come cheap or easily.

There is good reason to incorporate AI into a military force. Not only could it be capable of making decision much faster and thereby responding faster to a threat it would be on duty 24/7/365. No bathroom breaks for them. AI would be capable of analyzing military intelligence like satellite photos much faster and with less margin for error making human decisions more accurate and better informed. On the battlefield AI could be incorporated into weapons systems to help increase accuracy or even become its own weapons system. The next war could be fought by AI. One does have to wonder if AI can explain itself and the decisions that it makes could it be court-martialed?

Comments are closed.

Scroll to Top