What we call AI today isn’t really artificial intelligence. When you have a conversation with an AI chat bot you’re not talking to another thinking entity, you’re interacting with software that has been designed to give the illusion of conversing with another intelligent being. The technology has advanced enough that the illusion can be very convincing, but it is still only an illusion. That’s why I don’t fear LLMs being self aware and taking over the world, because they’re not real intelligences. They don’t have the ability to think for themselves because they don’t have the ability to think.
Edit: please read ricecake’s reply for an important correction to my comment.
Your conclusion is correct, but your terminology is wrong.
What we call AI today is AI, because AI doesn’t mean “capable of thought”, consciousness, sapience or anything like that.
It’s capable of producing a coherent output adapted to observed circumstances. That’s roughly as far as the notion of intelligence goes, and it’s a very low bar. You don’t need a lot of intelligence to be intelligent.
The people who coined the term were interested in how you make computers react to their inputs dynamically instead of acting closer to what we might now think of as a saved macro.
“It’s intelligent because rather than comparing against a list of every known typo, it sees it’s not a word in its list, and then replaces it with the one requiring the fewest edits to reach. It learns by adding your corrections to the known word list.”
It’s fallacious to say that because they don’t “think” that they can’t behave exactly the same as thinking things would, including the appearance of self-awareness and taking over the world.
What we call AI today isn’t really artificial intelligence. When you have a conversation with an AI chat bot you’re not talking to another thinking entity, you’re interacting with software that has been designed to give the illusion of conversing with another intelligent being. The technology has advanced enough that the illusion can be very convincing, but it is still only an illusion. That’s why I don’t fear LLMs being self aware and taking over the world, because they’re not real intelligences. They don’t have the ability to think for themselves because they don’t have the ability to think.
Edit: please read ricecake’s reply for an important correction to my comment.
Your conclusion is correct, but your terminology is wrong.
What we call AI today is AI, because AI doesn’t mean “capable of thought”, consciousness, sapience or anything like that.
It’s capable of producing a coherent output adapted to observed circumstances. That’s roughly as far as the notion of intelligence goes, and it’s a very low bar. You don’t need a lot of intelligence to be intelligent.
The people who coined the term were interested in how you make computers react to their inputs dynamically instead of acting closer to what we might now think of as a saved macro.
“It’s intelligent because rather than comparing against a list of every known typo, it sees it’s not a word in its list, and then replaces it with the one requiring the fewest edits to reach. It learns by adding your corrections to the known word list.”
Thank you for the clarification.
What about “Artificial” do you not understand?
It’s the artificial (e.g. not natural/human) part that is all the programming and stuff making it do what it does.
Sounds like you try to refute that an AI is conscious or something.
It’s fallacious to say that because they don’t “think” that they can’t behave exactly the same as thinking things would, including the appearance of self-awareness and taking over the world.
terminator was not a documentary
No but it was a warning, just like 1984 or most sci-fi about AI
If you somehow took the film as a warning about computers generating pictures as directed by a human user you’re really stretching!
This is correct.