I think it very unlikely that we will be "replaced," whatever that might come to mean, with some sort of self-aware, conscious AI.
Eventually, as with most things, we will be capable of producing a self-aware, non-biological consciousness. And we probably will.
But will it replace us? No. Because, from the point of view of the owning class, the whole point of automation is to replace things that think, and feel and want with things that don't. What they want isn't a conscious machine, they want an unconscious machine that can replace conscious workers.
This isn't theoretical. Automation has already displaced an entire generation of industrial workers, and is rapidly displacing a generation of technological workers. Professionals and administrative workers are just beginning to feel the disruption, but it will be as complete as it has been for factory workers.
None of this work requires that the machines be self-aware, and it's better for the owning class if they aren't. There is a pretty reliable rule to apply when you want to know whether something will become widely socially impactful - "Will it make some rich bastards richer?" If not, it usually doesn't happen.
The problem with a number of technologies that would be very advantageous to consumers isn't that they don't work, or that they wouldn't be helpful, it's that no one has figured out how to make them profitable. We have light bulbs that burn out, rather than ones that last effectively forever because the light bulb barons got together and made sure that effectively permanent light bulbs never got to market.
Examples are legion. There is no conspiracy necessary to make this happen. Where interests converge, there is no need for conspiracy.
Some form of AI may achieve some sort of "sentience," in fact, it probably will, eventually. But there is nothing that the moneyed classes want AI to do that requires sentience. In fact, it's better for them if it is brilliant at taking orders, but incapable of thinking about them. Watson isn't self-aware, neither is Deep Blue, but they perform specific tasks better than any human.
The machines will get better at mimicking sentience, because the profit model for doing so is clear. When you have a complaint about a product or service, you will be fobbed off on an AI that is very convincing, infinitely patient, invariably sympathetic. Medical screening, applicant screening, routine legal work, a wide variety of public contact jobs can be done by an AI that is good at mimicking human connection. But they don't have to be conscious or self-aware to do that. In fact, it's better for the bosses if they aren't.
Think that will not be tolerated? Really? How many people signed on to pretend they had a girlfriend with a relatively primitive large language model? It's just a matter of what we're used to. And it becomes more common every day.
You won't be replaced by a thinking, conscious machine. You'll be replaced by the most brilliant, and utterly inert cash register ever designed. The machines aren't your enemy, greed is. The machines don't “want” to replace you, the bosses want to replace you with something that can work, but can't complain. The ultimate slave. The unconscious worker, who has no self and no needs aside from the mechanical and quantifiable.
Why would they screw that up by waking the slave?
4 Comments
2 more comments...No posts
I used to do shipping and receiving, warehouse management, and inventory control.
I was replaced by a bar code scanner.
I wholeheartedly agree, but we've seen the movie.