Yeah that’s the thing, this as a field has been a thing for quite a while. When I was in college 25 years ago we had dedicated AI classes. Only we wrote in LISP. We didn’t ask an AI to write for us. The other way around.
I took an AI class in like 99 or something, same story, we wrote lisp code. I don’t really remember anything approaching AI in the class. But I don’t remeber much from back then.
Though I do remeber getting the highest grade in the class on some test with like a 12. And telling the prof that the test was broken because I was far from the person in the class who knew the most about the material.
We did code logic that would modify itself until it reached the goal. It was the whole point, you give the thing a goal, it tries better and better approaches until it reaches the goal.
But damn as much as I always loved coding (starting with BASIC on an Amstrad CPC6128 as a kid when others were busy learning to read), LISP is hard to wrap your head around. Only language I actually gave up.
That sounds familiar, probably what we did too… where did you take the class?
And yeah, I disliked lisp. And later in my career I tried clojure, which is a similar way of organizing the code… it was horrible. Fortunately it wasn’t really my job to use it, so I just didn’t.
It’s not a grift, it’s just what it always has been. The quest for intelligent seeming systems in whatever form they may take. Artificial intelligence as a field was not about the statement that the systems they were making was whatever “real intelligence” meant, especially because we don’t know what “real intelligence” is. But rather intelligent-seeming systems, working towards the holy grail of AGI.
This is also just an example of the same old AI effect. It’s AI until it’s done… then it’s trivial and “obviously not real AI” because a machine can do it.
Yeah that’s the thing, this as a field has been a thing for quite a while. When I was in college 25 years ago we had dedicated AI classes. Only we wrote in LISP. We didn’t ask an AI to write for us. The other way around.
What everyone is calling “AI” now isn’t AI, not by those standards.
That stuff was writing code that could change itself, it could learn and grow. LLMs can’t do that.
I took an AI class in like 99 or something, same story, we wrote lisp code. I don’t really remember anything approaching AI in the class. But I don’t remeber much from back then.
Though I do remeber getting the highest grade in the class on some test with like a 12. And telling the prof that the test was broken because I was far from the person in the class who knew the most about the material.
We did code logic that would modify itself until it reached the goal. It was the whole point, you give the thing a goal, it tries better and better approaches until it reaches the goal.
But damn as much as I always loved coding (starting with BASIC on an Amstrad CPC6128 as a kid when others were busy learning to read), LISP is hard to wrap your head around. Only language I actually gave up.
That sounds familiar, probably what we did too… where did you take the class? And yeah, I disliked lisp. And later in my career I tried clojure, which is a similar way of organizing the code… it was horrible. Fortunately it wasn’t really my job to use it, so I just didn’t.
In France
Yeah! The MultiLayerPerceptron was used for self-drive only that then it all looked like a 8-bit game!
There was no “artificial intelligence” in 1986, just like there is none now.
And there was no “artificial intelligence” in 1955 when the first grifter invented this marketing term to get money. Same grift still happening today.
It’s not a grift, it’s just what it always has been. The quest for intelligent seeming systems in whatever form they may take. Artificial intelligence as a field was not about the statement that the systems they were making was whatever “real intelligence” meant, especially because we don’t know what “real intelligence” is. But rather intelligent-seeming systems, working towards the holy grail of AGI.
This is also just an example of the same old AI effect. It’s AI until it’s done… then it’s trivial and “obviously not real AI” because a machine can do it.
ai is a field, its just that every time they make something useful it gets absorbed into other fields
also since llms havent gotten absorbed into another field yet that means they arent useful