Weaponised AI - Project Mavern

[Red Sky at Morning]

I was woken early by the noisy birds outside our home and doing a bit of random searching.

I came across Google's "Project Mavern". Having watched "Person of Interest" a couple of years ago, my ears pricked up...

Google will end Project Maven military contract in 2019

Google is ending its involvement with Project Maven, the controversial Pentagon research program that sought to use AI to improve object recognition in military drones.

Diane Greene, head of Google Cloud, told employees during a Friday meeting that the company will let its current contract with the Defense Department lapse in 2019 and that it will not pursue a new one, according to the New York Times and Gizmodo. The announcement comes shortly after Google said it would draft an ethics policy to guide its involvement in future military projects — one that would explicitly ban the use of artificial intelligence in weaponry.

“It is incumbent on us to show leadership [in the ethical use of AI],” Green reportedly said during the meeting.

Source

So what did I notice?

Despite having massive reservations about weaponising and abusing AI to erode freedom, Google are going to work on the project for another year! Cynically, one could argue that this is a PR piece by Google to wash its hands of the results of opening Pandoras Box and giving it to the military...

I might buy a baseball cap and fight back against the surveillance ;-)

 

[Helioform]

One of the people who protested against Maven was one of my CS professors (Yoshua Bengio) at the university of Montreal. He did some important work on neural nets which I thought at the time was only theoretical and probably would never be used widely.

He is working on artificial empathy theory for AI which he thinks is the only way an AI should be built so it cannot harm humans. Still, there is always the danger that someone could hack into the AI code and make it do awful stuff.