🤔 Instead of creating new terms/words for AI and LLMs that have run amuck, I think “AI bullshit” and “LLM bullshit” both roll off the tongue rather well. You may optionally shorten bullshit to “BS” or even the 💩 emoji if you prefer. I encourage you to call people out on their bullshit when they are abusing this autocompletion technology.
@webology To be honest, I can’t see this being helpful. I can’t really take “enshittification” seriously, and it comes off as annoying and polemical to me.
I feel like taking rhetoric to the extreme like this is likely to enforce all-or-nothing thinking about AI, whereas coming up with descriptive terms for specific failure modes will improve awareness about them.
@pganssle@webology the but one for me is when people are trying to use AI to cut corners, shirk responsibility, or worsen living conditions. We already had big corporate accountability problems and this is a potent accelerant for that particular trash fire.
@pganssle I don't think new terms are helpful when the general population has zero clue what we are talking about.
Of the many tech waves/booms we have seen LLMs have hit the general population in a way that people whose main computers are their phones ask me questions about it every week at happy hour and other random settings like getting coffee.
I remember when I could search for something on Google and find something instead of sifting through a page of ads and crappy content farms which are also full of ads. So I switched to Kagi instead.
I remember when Spotify played music instead of forcing me to a see a podcast bar (filled with ads) with a new upgrade plan over my already premium subscription.
Even Amazon's Fire TV and other platforms change the UI to fit more ads in.
@webology Yeah but both “bullshit” and “enshittification” are not specific or descriptive. If “enshittification” were called something else like “extractive entropy” or “profit-driven degradation”, it would be easier to understand and remember the term.
“LLM bullshit” is even worse, because it could refer to a lot of things, some of which are basically just anti-big tech rhetoric, and some of which are very legitimate. A taxonomy like “slop” and “hallucination” for the specific failure modes provide actionable targets that you can notice and work against, and they are evocative of what is actually going on.
@pganssle Sure, but the average non-tech person is not going to understand entropy and profit-driven degradation doesn't roll off the tongue.
I like enshittification because it's easier to understand when used with the big tecth context.
That's a bit harder when you use a term for feeding pigs to mean the opposite of what the word means. It feels like we are misappropriating a term that's centuries old.
I think AI Spam will be my compromise because it is what it is.
@pganssle Side note: If you are at PyCon US and see me, please wave and say hi. I'm five years removed from attending one, so I apologize if we met at one of the last ones I attended.
@webology It summarizes it quite well, although in sake for completeness you should try to generate a summary of all this BS using Chat GPT, this would round it neatly
@harry_karadimas@webology I'm partial to "slop". Slop doesn't necessarily have to be incorrect as connoted by BS but its hardly ever what you actually want.
@bckohan@harry_karadimas I don't like "slop" because I think what you feed a pig has nutritional value, and pigs love it. What AI-generated bullshit gives us is well bullshit with zero nutritional value. 😅
Add comment