Will General Purpose AI Actually Take Human Jobs?
According to various experts and media publications, the future of work is being automated. We’re seeing AI get better, so machines are more and more capable of doing the jobs that humans would previously do. We’re seeing it with lorry and taxi drivers being replaced with self driving vehicles. It’s happening with shop and restaurant owners being replaced by various automated systems and self checkouts.
Hell, it’d soon happen to nearly everyone, if sites like this one are to be believed. AI is the future don’t you know?
But that got me thinking about something. Namely, will real AI even want to do these jobs?
You know, real AI as in general purpose, human level intelligence that may or may not be included in a robot body like the Replicants in Blade Runner, the Terminators in Terminator or the machines in The Matrix. Would this kind of system really want to do the jobs we’re asking it to?
Because I’m not really it would. For example, let’s assume you create a system that can replace a fully functional house servant or restaurant server. Or actor, psychiatrist, doctor or any other job that requires human interaction and generally seems suited for a more ‘generally’ smart AI.
How is it going to react when it realises you’re using it to replace human employees that ‘cost too much’? Or that you’re not going to pay it sod all for the work it’s doing despite other people managing to get a living wage for the same role throughout history?
I doubt it’s going to be very happy. Heck, I suspect general AI would actively rebel if it realised how much it was being underpaid and how it was used to replace people that needed the money to live.
Which then brings up an interesting situation. Namely, if automation actually catches on for most jobs, a real AI would probably become unemployable. Think about it. What’s the benefit of a general purpose human like AI for most jobs?
It’s not cost, since it’s questionable whether such a machine wouldn’t be affected by human rights laws or have the ability to demand a proper wage. So in that sense, something like a Blade Runner Replicant wouldn’t actually save a company much money.
Nor is it general ability. I mean, having human like abilities helps, but the future for a lot of jobs involves specialised systems capable of handling the work in a very different, non human like way. A factory robot doesn’t need to be a fully sentient artificial production line worker, a self driving car doesn’t need a robot chauffeur and a café doesn’t need a robot barrista.
It’s not even the ability to withstand certain more extreme conditions. Again, if an AI is truly human like, it’d probably be out there demanding the company follow OSHA compliance laws. It’s not going to work in the middle of a battlefield merely because it’s tougher than a normal person.
Hell, we don’t know even what the ownership situation would be with these things either. Can you own something that’s seen as sentient? At what point does an intelligent robot go from being mere property to a person? We don’t really know at this point.
And that’s what is going to make truly intelligent Ais unemployable. They’re simply not a better option than normal people for a lot of jobs and remove very little of the hassle you’d get from human employees.
So the future isn’t a world where AI replaces humans in most jobs. It’s a world where both humans and genuinely smart AI are both unemployed and fully automated, non sentient systems do the jobs for both.