People on the Fediverse get angry because I don’t use AI as a synonym for “generative LLMs.”There’s a reason for that: it’s fucking stupid
-
People on the Fediverse get angry because I don’t use AI as a synonym for “generative LLMs.”
There’s a reason for that: it’s fucking stupid.
I’ve been playing video games my entire life. Half the reason I still use social media is to talk about video games.
And in games, AI has always meant enemy behaviour, pathfinding, squad logic, and all the little scripts that make NPCs either brilliant or braindead. Nothing to do with LLMs.
And this week I bought an LG OLED TV. The box blasts AI like as a feature headline. Is LG talking about generative models? Obviously not. It’s talking about DSP pipelines, scene classifiers, and super-scaling. Real, old-school applied ML.
What bothers me is this constant habit of redefining technical terms around whatever flavour-of-the-month happens to annoy you. If you’re mad about AI image slop, suddenly AI must only mean “that thing that generates slop.”
It’s the same pattern as people using algorithm to mean “the part of social media I don’t like” instead of the utterly basic concept it actually is.
I’m not interested in that kind of outrage-driven semantic drift. I’m not going to take broad technical terms and collapse them down to the single thing currently making someone angry.
-
People on the Fediverse get angry because I don’t use AI as a synonym for “generative LLMs.”
There’s a reason for that: it’s fucking stupid.
I’ve been playing video games my entire life. Half the reason I still use social media is to talk about video games.
And in games, AI has always meant enemy behaviour, pathfinding, squad logic, and all the little scripts that make NPCs either brilliant or braindead. Nothing to do with LLMs.
And this week I bought an LG OLED TV. The box blasts AI like as a feature headline. Is LG talking about generative models? Obviously not. It’s talking about DSP pipelines, scene classifiers, and super-scaling. Real, old-school applied ML.
What bothers me is this constant habit of redefining technical terms around whatever flavour-of-the-month happens to annoy you. If you’re mad about AI image slop, suddenly AI must only mean “that thing that generates slop.”
It’s the same pattern as people using algorithm to mean “the part of social media I don’t like” instead of the utterly basic concept it actually is.
I’m not interested in that kind of outrage-driven semantic drift. I’m not going to take broad technical terms and collapse them down to the single thing currently making someone angry.@atomicpoet Yeah, the term AI has been around since at least the 70s (possibly longer) and I've always been irritated by its lack of a precise definition. It doesn't have to be a bad thing, but most things that advertise themselves as "AI" these days are garbage.
See also: my irritation with not being able to use the word cryptography because everyone now thinks it means cryptocurrency rather than cryptography.
-
@atomicpoet Yeah, the term AI has been around since at least the 70s (possibly longer) and I've always been irritated by its lack of a precise definition. It doesn't have to be a bad thing, but most things that advertise themselves as "AI" these days are garbage.
See also: my irritation with not being able to use the word cryptography because everyone now thinks it means cryptocurrency rather than cryptography.
Jonathan Lamothe I simply don’t think we should always be re-defining technical terms to refer to a specific thing in the zeitgeist that fuels outrage and resentment.
-
People on the Fediverse get angry because I don’t use AI as a synonym for “generative LLMs.”
There’s a reason for that: it’s fucking stupid.
I’ve been playing video games my entire life. Half the reason I still use social media is to talk about video games.
And in games, AI has always meant enemy behaviour, pathfinding, squad logic, and all the little scripts that make NPCs either brilliant or braindead. Nothing to do with LLMs.
And this week I bought an LG OLED TV. The box blasts AI like as a feature headline. Is LG talking about generative models? Obviously not. It’s talking about DSP pipelines, scene classifiers, and super-scaling. Real, old-school applied ML.
What bothers me is this constant habit of redefining technical terms around whatever flavour-of-the-month happens to annoy you. If you’re mad about AI image slop, suddenly AI must only mean “that thing that generates slop.”
It’s the same pattern as people using algorithm to mean “the part of social media I don’t like” instead of the utterly basic concept it actually is.
I’m not interested in that kind of outrage-driven semantic drift. I’m not going to take broad technical terms and collapse them down to the single thing currently making someone angry.@atomicpoet there was a time when spellcheck was considered AI. It really is such a general term. Like "computer" actually.
-
@atomicpoet there was a time when spellcheck was considered AI. It really is such a general term. Like "computer" actually.
wolfkin I mean, Grammarly is considered AI. As is auto-complete.
