AI in the Military?

As a resident of the United States of America, I fear the militaries of other countries that are extremely competitive with ours. China, Russia, and North Korea are real threats to American sovereignty. Currently, people think the next game changer for national defense is with Artificial Intelligence (AI). An article from Technology Review stated, “Early this year, the US announced a grand strategy for harnessing artificial intelligence in many areas of the military, including intelligence analysis, decision-making, vehicle autonomy, logistics, and weaponry. The Department of Defense’s proposed $718 billion budget for 2020 allocates $927 million for AI and machine learning.” The former Secretary of Defense, Jim Mattis, sent a memo to President Trump stating that the military was falling behind when it came to matters of artificial intelligence. The Chinese and Russian governments have made their goals public to incorporate AI into their military technologies. 

Recently Chinese researchers have been testing how to find the faults within the Artificial Intelligence involved in the Tesla Autopilot features. The AI software that Tesla uses reacts very differently to situations in comparison to how the human brain would. The researchers began to understand how to make the AI see what they wanted it to see. They were soon able to deceive the Tesla into completing tasks such as driving into oncoming traffic and unnecessarily turning on the windshield wipers. It did not take very long for the researchers to study the Tesla and figure out how to modify the information that the AI was receiving. From this, it can be concluded that a growing concern is whether or not Artificial Intelligence should be used in the military forces. If the highly advanced AI in Tesla software can be altered with a bit of effort, what does that mean for the technology being used for the military forces? Should the United States continue spending extremely large portions of the government funding towards military Artificial Intelligence if there is a chance it could be dangerous? With constantly advancing technology, these are new matters that have to be taken into consideration.

Personally, I believe it may be of best interest to focus on utilizing government funding for other types of military powers. I do not think it is beneficial to potentially risk having software that can be readily corrupted as the main point of defense for the nation. I think more people are intrigued by the idea of having a bigger Artificial Intelligence presence in anything today, but that does not mean it is the best move to take. There is no doubt that AI in everyday items such as cars, household appliances, and smart devices are incredibly helpful and resourceful tools, but that does not mean that the technology is ready to be the main strategy for military use in the United States. If Tesla software can be altered and directed to driving into traffic, there is the possibility that military AI can be modified, too. It poses too much of a risk for the people in the country, and there are other approaches and tactics that the military can utilize instead.

https://www.technologyreview.com/s/614497/military-artificial-intelligence-can-be-easily-and-dangerously-fooled/