The future of artificial intelligence and how you can save our world.
There are only three things you need to know about artificial intelligence. First, it’s coming. Second, you can’t stop it. Third, it will be smarter than humans.
This is the central argument of ‘Scary Smart’ (Bluebird, £18.99, ISBN 9781529077186), an extended essay on AI’s ongoing journey to the ‘singularity’ that will take place in the middle of this century, when machines will be “a billion times more intelligent” than ourselves. If AI will become Einstein, we’ll be no more than flies. What, asks author Mo Gawdat, is going to stop ‘Einstein’ swatting the flies?
It’s a good question and one that Gawdat is well-positioned to address. As a former chief business officer at Google X, serial entrepreneur and start-up mentor, he’s native to a world that’s unafraid to take on big concept horizon scanning. The answer – and it’s hardly a plot spoiler because it’s printed on the cover as the book’s subtitle – is you. The reader can save the world from the impending AI Armageddon because the power is in our hands to change the course of history for the better. It’s a double-edged sword because technology has the potential for both good and evil. Remember, Gawdat says, “If we control AI, it won’t live up to our expectations and if we don’t, we risk it going rogue.”
Gawdat’s method of recapping with such easy-to-remember soundbites helps to move his argument along at an exhilarating pace. Sometimes there’s a lack of technical ballast to steady his position, but since Scary Smart is a work of conjectural futurology, then perhaps it’s enough to take his former position at Google as all the credibility we need. There are still times when we’re left wanting more guidance than a pull quote exhorting us to remember that “a good machine in the wrong hands is a bad machine”, or “it’s not clear whose interests the machines will have at heart’, or even the manifestly debatable “intelligent machines will feel more emotions than we can ever feel.” While this approach is a superb vehicle for raising the ethical questions at the core of the future of AI, it is perhaps less good at answering them.
As worthy as his intentions are – and as fascinating as the author’s approach to his discussion of the future of AI might be – ‘Scary Smart’ will inevitably test the patience of the more technologically literate audience. Maybe it’s the harmless, upbeat Google linguistics of the TED-talk generation, but when an author describes his insights as ‘unique’ not once but twice before the end of the third page, I can see a more conservative readership cynically preparing themselves for the exact opposite.
Then again, the style of ‘Scary Smart’ is self-consciously informal because, as Gawdat tells us in the passage detailing the evolution of human intelligence, “spoken and written language in words and maths” is a “killer app.”
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.