It seems that all AI can do, it can just "replace the average". Write an average quality software, do good enough business analysis, prepare mediocre report, do a bearable, but not so great translation, give you some basic medical advice, which is accurate 80% of time (but what if you are not lucky to have typical conditions?). It can also solve average difficulty level problems in math or physics.
That what it cannot do is to invent new drugs, cannot figure out cheap energy sources, find something better than Standard Model in physics, and anything that would contribute to overall human progress.
The result is that it can replace a lot of jobs, it will efficiently killed any research done and the level of master/PhDs, because majority of those can be done now with the help of AI (we already see this, the burst in new scientific articles - nothing groundbreaking, but publishable).
As a results we will have soon a gap, as only very determined people would go into science, analytical jobs, etc.
In one way, this is OK, we had too much science research that was done only for the sake of producing more "papers", we also had a lot of people in the companies doing "bullshit jobs".
But on the other hand, we might face soon situation in which there will be almost no people in Academia, so who will teach people, in addition, science worked often by doing wrong, unreasonable, crazy things, testing vague ideas and come up with something new and great. This required a lot of people doing science without any good results coming out, a lot of seemingly wasted effort just to hit several great ideas (a bit like Venture Capitals work).
This might also lead to decreasing number of white collar jobs. We can have a new "rust belt", the one for blue collar fall happened because of jobs offshoring, now we are "offshoring" white collar jobs to a few companies who offer AI models.
This all will affect badly society without any huge, visible benefits for it as a whole.
That what it cannot do is to invent new drugs, cannot figure out cheap energy sources, find something better than Standard Model in physics, and anything that would contribute to overall human progress.
The result is that it can replace a lot of jobs, it will efficiently killed any research done and the level of master/PhDs, because majority of those can be done now with the help of AI (we already see this, the burst in new scientific articles - nothing groundbreaking, but publishable).
As a results we will have soon a gap, as only very determined people would go into science, analytical jobs, etc.
In one way, this is OK, we had too much science research that was done only for the sake of producing more "papers", we also had a lot of people in the companies doing "bullshit jobs".
But on the other hand, we might face soon situation in which there will be almost no people in Academia, so who will teach people, in addition, science worked often by doing wrong, unreasonable, crazy things, testing vague ideas and come up with something new and great. This required a lot of people doing science without any good results coming out, a lot of seemingly wasted effort just to hit several great ideas (a bit like Venture Capitals work).
This might also lead to decreasing number of white collar jobs. We can have a new "rust belt", the one for blue collar fall happened because of jobs offshoring, now we are "offshoring" white collar jobs to a few companies who offer AI models.
This all will affect badly society without any huge, visible benefits for it as a whole.