This editorial was first published in The Providence (Rhode Island) Journal, a fellow GateHouse Media publication. Guest editorials don't necessarily reflect the Daily Messenger's opinions.
Hungry, you come home to an empty house and speak a sentence. You awaken a receiver-speaker that hears you, converts your words into code and, in a brief moment, summons a pizza delivery boy carrying a 12-inch pepperoni.
Or: A drone scans the shelves at a warehouse, observes that toilet paper is in short supply and orders another pallet. No human is involved.
Or: You buckle yourself into a driverless car and catch up on your email while you’re transported to the office. Along the way, you never look up to gaze out the window.
Smart devices are beginning to change the way we live. An intelligent device has the potential to make your life simpler, saving you time and effort while expanding your ability to get things done.
But some scientific thinkers are beginning to wonder if we should worry about this trend.
What happens when machines surpass us in intelligence, foreseeing situations that we can’t, anticipating and responding to them before we know what’s happening? What happens when a system refuses to turn itself off? What happens, in other words, when the robots take over?
This question has moved from the realm of science fiction to one that is seriously discussed by such accomplished people as Bill Gates, Elon Musk and Stephen Hawking. Mr. Musk, in fact, is among those who have committed more than $1 billion to OpenAI, an “altruistic” nonprofit devoted to ensuring that a deep, human-level artificial intelligence system works for the benefit of people, not against them.
These people see the approach of what Ray Kurzweil and others call “the Singularity,” when human intelligence is meshed with and surpassed by machines to produce unexpected, real-world results.
At that time, machines will be fully capable of “deep learning,” of sifting data to recognize patterns we don’t see, and improving themselves autonomously.
As the abilities of these systems grow and express themselves, we will no longer be in control. But unlike parents who raise children who become smarter, richer and more accomplished than they are, we can’t easily teach moral values to a super-intelligent system, especially since human programmers are often not all that hot on moral values themselves.
As Stanley Kubrick posited in his 1968 movie “2001: A Space Odyssey,” a computer with super-intelligence might well develop the inclination to take charge itself for its own selfish reasons.
For many smart people, this is a frightening vision. They imagine a world in which humans are secondary to a network of super-intelligent systems. In such a world, humans could be reduced to being a colony of ants, or worse. As Oxford University professor Stuart Armstrong puts it, if things go wrong, artificial intelligence poses “an extinction risk” to humans.
We already know that computer problems can snarl credit scores, turn off power grids and shut down airlines. Imagine when these systems are self-aware and fully interdependent, altering our lives in unforeseen ways.
That’s why we’re increasingly seeing expressions of concern by academics and technology gurus. They worry that, while we’re rushing to create devices that do clever things such as driving cars, too few of us are taking the long view, looking down the road to make sure we humans are going where we want to.
Even in the midst of our busy lives, this is something to ponder.