Should we even call it AI?
Words matter, but sometimes more to rhetorical approaches than to any real understanding.
Yes.
Look, “artificial intelligence” is super vague. It can mean self-aware, reasoning constructed consciousness - Hal 9000 or Data the android. In typical usage, however, it’s used to mean something with simulated reasoning logic. The game industry has been using “AI” in this sense for decades.
Marketing around the current generation of AI tools has taken advantage of the ambiguity between common use of the term and popular, scifi-inflected expectations. Large language models — the specific type of AI people tend to be talking about now — are not self-aware consciouses. They don’t actually think1. If friends and family seem confused about this, try to dissuade them of that notion.
Just insisting we not call LLMs “AI” every time someone uses the word, though, is language-policing. It derails a conversation or changes the subject.
Footnotes
-
Whether humans actually think is open to question, which does change the debate about LLMs somewhat. But bringing that up is even more of a distraction. ↩
Add a comment