I propose you eat less hype about the slop generators. AI doesn’t exist, and there is no reason to believe that we’re closer to understanding if it’s even possible. Machine learning algorithms have their uses and are used already a lot, and nobody is against that, but that’s not AI. LLMs being pushed everywhere, and it’s never useful or particularly liked, and that’s not AI either. My hunch is that this bubble will pop, leaving an unpleasant odour behind, which we will have to deal with for years after, and then tech bros will come up with a new bullshit that revolutionises the world and disrupts the universe, because there is no meritocracy and the world is stupid.
Something that has actual intelligence, or at least significant portion of building blocks of intelligence.
The problem, of course, is that intelligence is a complicated, complex, sprawling phenomenon, with no real ways to measure it as a whole, at least to any degree of reliability.
Learning, reasoning, critical thinking. Creativity, logic, problem solving, abstraction. Self-awareness, self-reflection, general sense of self. You need most of the elements of most of the groups (and probably more that is also important but I am missing right now) in order to even begin to talk about possibility of intelligence. Then somehow we will need to solve the philosophical zombie problem, and I don’t envy the researchers who will have to do that, but that’s way later down the line.
What we have right now very demonstrably doesn’t have almost any of those. What we call machine learning can be called learning in a very specific and reductive way, and whatever emerged phenomenon we observe from that is it’s own beast, but intelligence it is not. All the other boxes are not ticked, and some, like creativity or critical thinking, are the opposite of ticked. It might lead to something in the future (personally I doubt it, but can’t rule out), it might just as likely be something else, or nothing entirely.
I am very unsecure in my speculations on it, but those who have the most robust and optimistic answers right now are actually those who want to sell you something, and most of them are salespeople with the expertise of sales and nothing more.
I propose you eat less hype about the slop generators. AI doesn’t exist, and there is no reason to believe that we’re closer to understanding if it’s even possible. Machine learning algorithms have their uses and are used already a lot, and nobody is against that, but that’s not AI. LLMs being pushed everywhere, and it’s never useful or particularly liked, and that’s not AI either. My hunch is that this bubble will pop, leaving an unpleasant odour behind, which we will have to deal with for years after, and then tech bros will come up with a new bullshit that revolutionises the world and disrupts the universe, because there is no meritocracy and the world is stupid.
What constitutes AI by your definition?
Something that has actual intelligence, or at least significant portion of building blocks of intelligence.
The problem, of course, is that intelligence is a complicated, complex, sprawling phenomenon, with no real ways to measure it as a whole, at least to any degree of reliability.
Learning, reasoning, critical thinking. Creativity, logic, problem solving, abstraction. Self-awareness, self-reflection, general sense of self. You need most of the elements of most of the groups (and probably more that is also important but I am missing right now) in order to even begin to talk about possibility of intelligence. Then somehow we will need to solve the philosophical zombie problem, and I don’t envy the researchers who will have to do that, but that’s way later down the line.
What we have right now very demonstrably doesn’t have almost any of those. What we call machine learning can be called learning in a very specific and reductive way, and whatever emerged phenomenon we observe from that is it’s own beast, but intelligence it is not. All the other boxes are not ticked, and some, like creativity or critical thinking, are the opposite of ticked. It might lead to something in the future (personally I doubt it, but can’t rule out), it might just as likely be something else, or nothing entirely.
I am very unsecure in my speculations on it, but those who have the most robust and optimistic answers right now are actually those who want to sell you something, and most of them are salespeople with the expertise of sales and nothing more.