I'm not sure exactly. As people say, it just always happens. The only thing I think these people are wrong about is that this trend continues after true AGI. That's when all trends and models of the economy and everything else breaks down.

If I had to guess, I think a lot of people will just be moved into places where AI is not yet fully capable. Mass collective data training. The more people on it, the faster we'll get to true AGI. If the AI is not yet at true AGI, then that means there is obviously areas where it needs to learn.

The economic value of training AI, once it has a full capacity to learn well from training, will just be absolutely massive. So it will require work from home-based solutions to get more people into these areas and very quick turnaround and retraining of people into new areas. The economic value will certainly be there though to facilitate such a system.

I think we'll inevitably also start to see a greater amount of real-world value being created too. So there will be a large increase in real-life activity needing to be done, whilst the robotics side of things lags behind.

I think robotics will still lag behind for a while, even after true AGI is created. It will take some time to manufacture and deploy all the necessary robots to replace workers. So there will still be a lot of people with jobs safter the inception of AGI even but then, slowly but surely, they'll start to get replaced. Starting with the higher salaried jobs first, then down towards the minimum wage workers eventually.

I think there will still be physical work after AGI. But it will be incredibly low-paid and optional. Humans are still useful, but they'll just have to accept not being paid much at all i  order to stay economically viable against a robot.

No way to know, imagine telling someone from the Middle Ages influencer, machine learning engineer, etc would be jobs, there’s no way to know.