Just Like Mother Used to Simulate

By Gordon Hopkins Nobel Prize-winning Geoffrey Everest Hinton, a British-Canadian computer scientist, cognitive scientist, former vice-president of Google and cognitive psychologist, is referred to by many as “the Godfather of AI” due to his work on artificial neural networks. Surely that’s the guy to ask, if anyone is.Hinton is on record as saying the odds of humanity being wiped out by AI is somewhere in the area of 5 to 10 percent.Those odds might be worth a shot if you were thinking about dropping a few bucks on the Superbowl. When talking about the survival of the species, on the other hand, I think I might like to hedge my bets a little bit.Apparently, Hinton agrees. I’m always reassured when a smart guy agrees with me. Hinton said in an interview earlier this year, “The hope is that if enough smart people do enough research with enough resources, we’ll figure out a way to build them so they’ll never want to harm us.”I hope so, too. But how? Per Hinton the only real option is to give AI something akin to “maternal instinct.” Somehow make AI want to take care of humanity. That is because, ultimately, AI is going to become smarter than humans. AI already is smarter than some humans, although it is worth keeping in mind some humans are really, really dumb, so that bar was never set too high.Hinton explained, “The right model is the only model we have of a more intelligent thing being controlled by a less intelligent thing, which is a mother being controlled by her baby.”I understand the basic concept, but I do foresee one minor problem. This solution, making AI care about humanity, requires that the people behind AI, controlling the development AI, have to care about humanity.Does that sound like Elon Musk? Mark Zuckerberg? Jeff Bezos? All the billionaire “tech bros” pushing AI into every facet of our lives? If we take Hilton at his word, the future of humanity is in the hands of people who would sell humanity out for a nice car or a pound of butter.Interestingly enough, Professor Fei-Fei Li, an AI pioneer often known as the “godmother of AI” doesn’t entirely agree with Hinton. In response to Hinton’s remarks, Li said, “I think that’s the wrong way to frame it.”However, Li is calling for “human-centered AI that preserves human dignity and human agency.”Again, how can you expect concern for “human dignity” from people whose work is being used to fool humanity with fake photos and videos of politicians and celebrities. In reality, the solution being proposed now is the same Isaac Asimov wrote about nearly a century ago. “The Three Laws of Robotics” was just another way to assure the population their creations would never go rogue and hurt them.Even as a school kid, secretly reading “I, Robot” under my desk during history class because I could not force myself to care about the Louisiana Purchase, even then I understood the basic flaw in the system Asimov worked out: what if someone built a robot that DIDN’T follow the three laws?Of course, as a kid, I was worried about robot killers. I wasn’t worried about robots raiding my bank account or stealing my intellectual property and passing it off as original or lying to me about politicians.The real problem can be summed up like this: We cannot expect the makers of AI to create an AI that won’t hurt people because there is too much profit in making an AI that CAN hurt people.

Lois Wrigley

Wymore Man Arrested for Disturbing the Peace and Deadly Weapon

Related Articles

Are Marshmallow Peeps the New Pumpkin Spice?

Again

Time Marches On (But I Don’t)

The Year of the Fire Horse