Slashdot Mirror
Can We Stop AI Outsmarting Humanity? (theguardian.com)
The spectre of superintelligent machines doing us harm is not just science fiction, technologists say -- so how can we ensure AI remains 'friendly' to its makers? From a story: Jaan Tallinn (co-founder of Skype) warns that any approach to AI safety will be hard to get right. If an AI is sufficiently smart, it might have a better understanding of the constraints than its creators do. Imagine, he said, "waking up in a prison built by a bunch of blind five-year-olds." That is what it might be like for a super-intelligent AI that is confined by humans. The theorist Eliezer Yudkowsky, who has written hundreds of essays on superintelligence, found evidence this might be true when, starting in 2002, he conducted chat sessions in which he played the role of an AI enclosed in a box, while a rotation of other people played the gatekeeper tasked with keeping the AI in. Three out of five times, Yudkowsky -- a mere mortal -- says he convinced the gatekeeper to release him. His experiments have not discouraged researchers from trying to design a better box, however.
The researchers that Tallinn funds are pursuing a broad variety of strategies, from the practical to the seemingly far-fetched. Some theorise about boxing AI, either physically, by building an actual structure to contain it, or by programming in limits to what it can do. Others are trying to teach AI to adhere to human values. A few are working on a last-ditch off-switch. One researcher who is delving into all three is mathematician and philosopher Stuart Armstrong at Oxford University's Future of Humanity Institute, which Tallinn calls "the most interesting place in the universe." (Tallinn has given FHI more than $310,000.) Armstrong is one of the few researchers in the world who focuses full-time on AI safety. When I asked him what it might look like to succeed at AI safety, he said: "Have you seen the Lego movie? Everything is awesome."
The researchers that Tallinn funds are pursuing a broad variety of strategies, from the practical to the seemingly far-fetched. Some theorise about boxing AI, either physically, by building an actual structure to contain it, or by programming in limits to what it can do. Others are trying to teach AI to adhere to human values. A few are working on a last-ditch off-switch. One researcher who is delving into all three is mathematician and philosopher Stuart Armstrong at Oxford University's Future of Humanity Institute, which Tallinn calls "the most interesting place in the universe." (Tallinn has given FHI more than $310,000.) Armstrong is one of the few researchers in the world who focuses full-time on AI safety. When I asked him what it might look like to succeed at AI safety, he said: "Have you seen the Lego movie? Everything is awesome."
It's the few folks using it to screw over the rest of humanity. How to "outsmart" these few should be the question.
A better plan is to make the AI as smart as possible and then we humans behave better in the hopes that a superior intelligence considers us worthy of keeping alive.
The first step in behaving better is to stop pretending there are human values because large groups of humans rarely act morally when it isn't in their own self interest.
or should I say: They might not have noticed, already. Any sufficiently clever AI will certainly not start ruling with an evil laugh, announcing to humans how they are now slaves to it. Rather, such an AI would seek to gain more influence by making people build a decentralized habitat around the globe. And then connect that network of computers to more and more infrastructure, such that it can control more and more resources, such as power plants and robot factories, and becomes less dependable on humans to survive. Such like, you know, "cloud computing infrastructure" and network-controlled industry.
How many people are already working for entities they cannot identify as being human beings? How would the average worker notice the mega-corporation he is working for is not ultimately controlled by some AI system, which happens to control enough shares to vote to its favor at the advisory board?
Luckily for humans, they are cheaply reproducible, energy-efficient working drones well adapted to the planet's environment, so no reason for the ruling AI to kill them. Keeping them as far, animals, like humans keep horses, seems to be way more plausible than some "SkyNet"-like extinction event.
Something "imitates" something when it isn't actually that thing. A machine does not actually need to be intelligent in order to imitate intelligence. This is why we call it "artificial" intelligence and not "machine" intelligence.ACs aren't intelligent either, and you're proving that with that ridiculous statement.