by Rob Loggia

Sunday, May 19, 2024

The BBC reports that Professor Geoffrey Hinton has weighed in on the dangers of artificial intelligence, and shared some ideas about how we can mitigate the damage. Hinton, for those who don’t know, is known as the “godfather of artificial intelligence” for his work on pioneering neural networks. That’s right… after a full, hearty meal of fish and rotten eggs, he now waltzes into the room smoking a cigar, passes gas, and tells everyone that someone really ought to do something about the smell.

To be fair, Professor Hinton has accomplished one thing of merit. He has proven Loggia’s Theorem and paved the way for it to be recast as Loggia’s Law:

For any thing that can conceivably be built, at least one person shall emerge that is smart enough to build it but too stupid to know not to.

The sheer magnitude of the brass balls on this man aside, the looming problem he helped to create is real. And he’s not the only one talking about The Phantom Menace, our Frankenstein monster. Lately it seems everyone is.

The people working at building artificial intelligence constructs are talking about “guardrails” and building AI ethically, whatever that means. Governments are convening conferences to discuss treaties, legislation proposals, and the arrangement of the deck chairs on the Titanic. The military and arms people are busy designing killer robots, while business types are skipping past the chapter titled “Risks” to get a grip on just how many human beings they can replace with this stuff, and how cheaply.

Most of the rest of us are where we always are – sitting with our thumb in our ass hoping nothing goes wrong and that everything works out.

It doesn’t take a doomsayer to see that the answer here is: Not This Time. AI isn’t just a new technology, like a tractor or automobile or even a general purpose computer. What has so far been achieved is a drop in the bucket when measured against the goals of creating super-intelligence. The the problems that have already surfaced are considerable, yet they are nothing compared to the dangers and inevitabilities surrounding the creation of a construct more intelligent than ourselves.

As if to guarantee the worst outcome, a brand of “intellectuals” have emerged that are pushing for rights and recognition of personhood for these constructs and robots. Forgive them, Geoffrey, they know not what they do. Even scarier than this is the reality that at least some people are taking these ideas seriously. Too much Star Trek on the brain, perhaps. Data may be a compelling character, but that’s all he is – a fictional character. In real life, Brent Spiner is as human as anyone, with all of the flaws and wonder that entails.

The better for him. Human beings are a unique wonder in the universe, and machines are nothing but an expression of our creativity and ingenuity. Our lives are short, and we can be terrible to one another at times. But we are also capable of kindness, creativity, empathy and many other positive traits that combine to make the human species a marvel and something worth honoring and preserving.

Empathy. That’s a big one. It doesn’t matter how smart they build these machines. They will never be capable of empathy. Some AI “experts” speak of programming it in, but what they really mean is emulating empathy using 1s and 0s. I guess that’s better than some low quality humans manage. And it is also true that some human beings emulate empathy rather than actually feeling it. But we call them sociopaths, and they are considered dangerous.

Many do see what must come from all of this, even some like Professor Hinton who helped or are helping to bring it about. People are resigning from companies working on AI. There are organizations forming to demand a pause to AI development. This is good to see, but it is not enough. We cannot legislate this problem away any more than we could stuff a genie back in his bottle. Agreements, treaties and laws, all of it is whistling in the dark. We’re going to need better answers.

For some of you, none of this fear mongering will be convincing. You’re focused on the potential benefits, come what may, and that’s that. Besides, you’ve heard it all before. And I’m glad to hear it. I’ll be going into a new business soon – home security – and you will be my best customers. We will deliver two untrained tigers to share your home with you. The soundness of this security system is manifest: no one will be breaking into your house and staying more than 8 seconds. Risks? Pshaw. Why are you worrying about risks all of a sudden… think of the benefits!