Artificial intelligence is hot. Self-driving cars and autonomous robots are here. We converse with Siri and Alexa and Google Assistant. It’s easy to believe that these systems and devices have achieved some degree of human intelligence, perhaps that of a child. They seem almost alive.
It’s good to be reminded that they aren’t.
They aren’t to be trusted, not because they have evil intentions or faulty logic, but because they are still just dumb, program-driven machines. Ask the man who’s Tesla attempted to drive UNDER the trailer of an 18-wheeler. (Oh wait – you can’t, because he was decapitated.) Other tragic, more recent examples: the Boeing 737 Max 8 crashes. It appears that the cause was Boeing’s “Maneuvering Characteristics Augmentation System, or MCAS, which is designed to automatically command a plane down if it senses an imminent stall.” This is auto-pilot software that is supposed to be smarter and faster than human pilots. And yet, it relies on input from a single angle-of-attack sensor. Boeing’s solution? Feed the software additional input from a second sensor. And if both sensors are bad?
You can update the software and plug in additional sensors, but it’s still not intelligent , artificially or otherwise. No system controlling an airplane, under any circumstances, should nose dive into the ground or ocean at 400 mph. Common sense, right? That’s what AI is missing. It ain’t human, not even close. It’s not your friend. It doesn’t care if you live or die. Which is why it’s a bit unnerving to see it proliferating throughout technology we depend on every day.
As for me, I will continue to drive my car myself. Even at the risk of missing a text or Facebook post while I’m at it. At least until AI gets smarter than me. Who could have believed that would be so difficult?