The problem is that technology is a type of prisoner's dilemma writ large. AI is only going to get better, that's a fact. It might take a while to get there but it'll get there. Especially since the best description for one of humanity's facets is 'intelligently lazy' aka 'using that intelligence to make being lazy easier'.
To deny that means we're denying ourselves.
So, pragmatically, we must accept that, eventually, people are going to be out of work, permanently (MIT did the math on this back in the 20-teens, BTW, and they discovered that job retention rates have been decreasing due to automation since the 1970s, and it has been slowly accelerating). We could have set up things for a Jetsons-style future, but we didn't and thus in this position.
None, I'm just stating the obvious and the reality of things. Humans will always look to make better tools to make their comfort easier. What is being called AI is just another symptom of that.
The upper limit has not been discovered. Ai is a buzzword. Its all just math at scale. The same math we have used for many useful things, its just now we have the scale to do some harm other than a social media skinnerbox.
It can and will get more capable. The bubble can and will burst. Those two things do not mean that this technology will not be used, nor does it mean there cannot exist a version of it that is moral.
That is hypothetical. The real problem is higher up. a neural net isnt the problem, the problem is unteathered and unrestricted immoral data collection and misuse. This happens independently of some learned weights and biases.
Its good that people are talking about this and are angry. It would be better if these discussions were more focused on the source of the problem though.
To deny that means we're denying ourselves.
So, pragmatically, we must accept that, eventually, people are going to be out of work, permanently (MIT did the math on this back in the 20-teens, BTW, and they discovered that job retention rates have been decreasing due to automation since the 1970s, and it has been slowly accelerating). We could have set up things for a Jetsons-style future, but we didn't and thus in this position.
The impending ai economic bubble bursting:
A bubble bursting is only going to slow it down.
It can and will get more capable. The bubble can and will burst. Those two things do not mean that this technology will not be used, nor does it mean there cannot exist a version of it that is moral.
That is hypothetical. The real problem is higher up. a neural net isnt the problem, the problem is unteathered and unrestricted immoral data collection and misuse. This happens independently of some learned weights and biases.
Its good that people are talking about this and are angry. It would be better if these discussions were more focused on the source of the problem though.