I recently read a newly published book on artificial intelligence. One thing was said early in the book that struck me as important. Children learn languages by shared experiences with their caregivers. So a child learns words and sentences by sharing these with someone who already knows the language. And that means that, once learned, the words and sentences relate to the shared experiences, not just the words themselves. All of the human part of learning a language is created by those shared experiences. And that is something a language model, no matter how many trillions of words it reads, will never replicate. It can never share an experience. It is just studying language for likely responses to prompts, and because of the advanced capabilities of chips and data centers it can go through billions of possible outcomes in almost an instant. But it doesn’t have any idea of what it’s talking about, no real knowledge is behind it. Only possible outcomes, or probable outcomes.
Over the last few decades, technology has managed to put a screen between us for most of the work we do. We don’t drive to job sites to see the work being done. We have the worker send us photos. Or we have drones fly overhead to give us an aerial view. We chat back and forth, or send emails, two dimensional communication instead of the vast influx of information we receive when in person. Now, in addition to a screen between us, they want to add a robot. So now a screen and a robot between each person and any other person. They’re trying very, very hard to convince us that these robots are just like us. ‘Robots are people too!’ They will never be people. They are not capable of having shared experiences. They don’t share anything with anyone, except prioritized responses to queries sifted from vast amounts of text. Humans are so much more than that, which you and I know, but which, sadly, the tech bros have never picked up on. They don’t have all of the human qualities which come into play between people. We think they’re feeling, caring beings because we hear them speak words which, coming from a human would convey those qualities. But there’s no shared experiences to back that up. It’s a mimic, alien to humanity.
So the dream of the tech bros to never have to deal with (stupid) people again is to create substitutes for real humans who can do the work of real people. Every CEO’s fantasy can come true. No more workers to try to make happy, pay a decent wage, or provide incentives for. Alien intelligence can do it all with no opinions, no down time, no family issues and no illness. Just the other day I read an article in which the CEO (a billionaire) said AI works 24/7, it doesn’t take a lunch break and doesn’t require health care. To his way of thinking, it’s the perfect employee.
Let’s not be fooled by all the hype around these programs. In some specific areas they may be able to do the work of humans without the risk of fatal mistakes. Even then it may require oversight. But it would be very limited and not profitable enough to justify the many, many billions being spent on them, unless, unless they replace employees on a mass scale. Is this the goal of big corporations now? Have we entered an era where ordinary people are no longer needed or wanted, where the joy of working together as a team to accomplish goals will be replaced by 24/7 machines? I don’t remember signing up for that. I suspect the vast majority of people don’t remember it either.