We know the United States Government has plan ‘CONOP 8888‘ in place as preparation for the zombie apocalypse should there ever be one, however it seems like Google may be taking the lead on preventing any robot uprising should there ever be one.
Recently, documents were released online highlighting the work of Google’s ‘Deep Mind‘ team who’s working to keep the world from seeing a SkyNet situation where robots rise to enslave the human race. The primary goal is developing a ‘kill switch’ framework from preventing AI to interrupt human intervention.
Here’s what Timothy J. Sepal of Engadget writes:
Essentially, Deep Mind has developed a framework that’ll keep AI from learning how to prevent — or induce — human interruption of whatever it’s doing. The team responsible for toppling a word Go champion hypothesized a situation where a robot was working in a warehouse, sorting boxes or going outside to bring more boxes in.
The latter is considered more important, so the researchers would give the robot a bigger reward for doing so. But human intervention to prevent damage is needed because it rains pretty frequently here. That alters the task for the robot, making it want to stay out of the rain, and then adopting the human interruption as part of the task rather than being a one-off thing.
Researchers further mention:
“Safe interruptibility can be useful to take control of a robot that is misbehaving and may lead to irreversible consequences, or to take it out of a delicate situation, or even to temporarily use it to achieve a task it did not learn to perform or would not necessarily receive rewards for this,”
Unfortunately Deep Mind can’t confirm whether or not this ‘kill switch’ can’t be introduced to all algorithms such as those whose primary function is policy-search robotics as Sepal writes. Whatever that means. So… it seems like the human race is still prone to that dreaded robot apocalypse.
Where’s Sarah Connor when you need her?