Legendary astrophysicist, author and cosmologist Stephen Hawking has written an article that warns about the dangers of artificial intelligence,(AI). He looks at the way technology is moving forward, and what predictions have been made about the future. The article was part of a paper that Hawking co-wrote with fellow science professors Stuart Russell a computer-science professor at Berkley University and physics professors Frank Wilczek and Max Tegmark of the Massachusetts Institute of Technology.
Stephen Hawking was inspired to write the piece after watching Transcendence, starring Johnny Depp and Morgan Freeman. The film, which is in cinemas at the moment, looks at two opposing possible futures for humanity. One is the road in which AI is a strong and crucial part of our existence and taking over many aspects of human life. The other is an anti-technology perspective. However, Hawking warns about dismissing this sort of artificial intelligence simply as science fiction.
There have already been a number of technological advancements towards the idea of AI. Listed in the article are so-called “digital personal assistants” Siri, Microsoft’s Cortana and Google Now. Hawking also mentions self-driving cars and the computer that won Jeopardy!. These progressions are considered “pale” by the astrophysicist, who states that there could be many other kinds of advancements in the future.
Whilst Hawking writes that to successfully create AI would be one of the greatest achievements of the human race, he also sees potential problems in the future. The uses of AI are endless, world issues such as poverty, disease and war could become a thing of the past. But there is a price. Defence firms are already looking into how AI could be used to create weapons that are completely autonomous to eliminate their targets. Super-intelligent machines could self-replicate, improving on their faults and learning as they go.
As such, they could outsmart any human counterparts in the case of financial markets, invent better machines than humanity could come up with and manipulate human leaders to their own benefit. Hawking notes that in the short-term, issues with AI would simply be centered around who has the control of the technology, whereas the long-term problems could be whether or not it is able to be controlled. This again has forerunner in sci-fi, such as the Terminator films where humanity battles against machines with AI, ones that were once under the human domain, but who chose to rebel against human overlords.
In an effort to prevent technology from falling into the wrong hands, the UN and Human Rights Watch has suggested a treaty against such weapons, banning them from being produced. Hawking notes that the future cannot be predicted in terms of the changes in technology. He looks at the changes that have come about so far as evidence of this fact. While some elements of technological developments were predicted, Nikola Tesla predicted that people of the future would be able to send “wireless” messages and Isaac Asimov in 1988 predicted the internet, but whilst Asimov could see how “connected libraries” would allow the children of the future to have access to a seemingly endless field of technology, he could not foresee all the other elements that the internet gives us.
Stephen Hawking is calling the developments into artificial intelligence “the best or worst thing” that could occur to humanity in the future. He warns that not enough research is being devoted to the possible risks that are involved. As such, moving forward more research is needed to try to overcome potential problems before they arise, rather than waiting until it is too late to fix.
By Sara Watson