U.S military is going to make its first ever biggest investment in artificial intelligence (AI) systems, defense department’s research wing swore to spend up to $2 billion in U.S weaponry, in the coming 5 years.
Defense Advanced Research Projects Agency (DARPA) feels that it is their duty to solve military problems using new tech solutions. Tech rulers of Trump’s administration have also provided their strong support to the concept of injecting AI systems, since they want to be equipped well in comparison to Chinese and Russian military forces.
In July, payment of $885 million for the next 5 years was made to Booz Allen Hamilton, defense contractor, to start work on artificial intelligence programs. Further, payment of $93 million is expected to be made by 2019 to Project Maven, which is an object recognition based single biggest AI venture.
Earlier, Project Maven was being led by Google, but soon Google employees protested that they don’t want to work on a venture that will point out objects for military to assassinate. So, in June, company announced that it will no longer work on this software after their current contract ends.
Pentagon weapons systems have benefited greatly through Maven and other AI projects, they are now good at identifying targets and flying drones. However, till now, automatic computer-driven systems that are fatal to use have not been approved by the agency.
A Pentagon report predicted that in future, automatic computer-driven systems will be able to take action on the targets emitting dangerous forces.
“DoD does not currently have an autonomous weapon system that can search for, identify, track, select, and engage targets independent of a human operator’s input,” was mentioned in a Pentagon’s report. And it further predicted that “technologies underpinning unmanned systems would make it possible to develop and deploy autonomous systems that could independently select and attack targets with lethal force,”
Commanders are doubtful about machine reasoning and they are not yet ready to leave everything on AI systems, since there is a chance that in battlefield a new can situation arise that is not previously seen by the machine, and in that case it will not be able to take the right decision. For now, an AI system will only tell the soldier the percentage of confidence level of the target it will select.
“What we’re trying to do with explainable AI is have the machine tell the human ‘here’s the answer, and here’s why I think this is the right answer’ and explain to the human being how it got to that answer,” Steven Walker, DARPA’s director, claimed.
Presently, 25 programs are working on AI projects in DARPA, but it is planning to invest more money and expand its programs through new AI Exploration Program, launched in July. This program will give funding up to $1 million for each research that will explore how AI systems can be trained to comprehend different environments.
Walker said that it is “critically important in a war fighting scenario” that AI platforms make decisions amidst of distractions and then justify them to their soldiers.