I don’t necessarily consider pathfinding AI, but I don’t like that reasoning. If ai coded a set of rules for a units behavior, if that behavior responds to different conditions, I’d consider it AI. Even though it is purely deterministic.
That would make any non-trivial piece of code AI. Like ffmpeg, for example, or a chess computer. Complicated tools, sure, but with a bit of effort you can predict how they will behave in a given circumstance. Meaning you could set a trap for the chess computer and it would walk right into it. Every single time. No learning occurs.
Until a few years ago AI was effectively synonymous with AGI, Artificial General Intelligence, which requires the software to be able to adapt to new situations and be able to solve even unknown problems with as few or fewer attempts than a human would need.
I don’t necessarily consider pathfinding AI, but I don’t like that reasoning. If ai coded a set of rules for a units behavior, if that behavior responds to different conditions, I’d consider it AI. Even though it is purely deterministic.
That would make any non-trivial piece of code AI. Like
ffmpeg, for example, or a chess computer. Complicated tools, sure, but with a bit of effort you can predict how they will behave in a given circumstance. Meaning you could set a trap for the chess computer and it would walk right into it. Every single time. No learning occurs.Until a few years ago AI was effectively synonymous with AGI, Artificial General Intelligence, which requires the software to be able to adapt to new situations and be able to solve even unknown problems with as few or fewer attempts than a human would need.