Sorry Elon Musk, the machines will not win

Weblog: Cyber expert Ryan Calo writes paper to demolish belief in looming AI apocalypse

Last Saturday tech billionaire and SpaceX chief executive Elon Musk tweeted a warning that there is "vastly more risk" associated with artificial intelligence than with North Korea along with an image bearing the words "In the end the machines will win". Musk and others in the public eye including theoretical physicist Stephen Hawking and philosopher Nick Bostrum have been vocal about AI risks many times in the past, but cyber and robotics law expert Ryan Calo, faculty co-director at the University of Washington's Tech Policy Lab, begs to differ.

In an essay on AI policy, he points out that nothing in the current literature suggests that AI could model “the intelligence of a lower mammal in full, let alone human intelligence . . . This explains why claims of a pending AI apocalypse come almost exclusively from the ranks of individuals such as Musk, Hawking and Bostrom who possess no formal training in the field.” Ouch.

Finally, Calo says that even if AI was to reach the level of superintelligence (smarter than humans), there is nothing to suggest that it would then focus on world domination.

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3015350