Home | Notifications | New Note | Local | Federated | Search | Logout

Note Detail

Reply to @sun@shitposter.world
(eval 'Toast)@toast@donotsta.re (2026-08-06 16:52:36)
@sun I haven't really used it much? When I was new I made an account and posted answers (if you see any answers from deleted accounts with like 14k points there's a solid chance it's one of mine), but I quickly realized it's pointless.
The posters are not trying to help you learn something, they're pointfarming with the target audience being other pointfarmers. And even if they were trying to help you, it's just the blind leading the gullible.
Your post was a good insight though in that it's the same kind of failure category, just with a different modality, in that both SE and LLMs tend to produce poor quality answers with hidden faults that you need to already know about the thing to identify.
The difference is that LLMs will make things up wholesale or push towards sycophancy while SE will push towards "I know the answer but you haven't earned it" (they do not in fact know the answer).
---Reply---
Blurry Moon@sun@shitposter.world (2026-08-06 16:58:47)
@toast I don't want to sound like "prompt engineer" is a real job but asking questions to an LLM is in fact a skill and can produce wildly different results depending on the quality of the questioning. there is basically probably ten people on Earth that could synthesize concepts from both Fuschia and SEL4 OS and none of them have time to talk to me, but the LLM at all their source code and gave me actionable insight. Sometimes it's wrong, but the gains are way bigger than the downsides. you learn how to mitigate that too.

---Replies---

『 Fristi, enjoyer of Kosuzu 』@fristi@56k.dile-up.nl (2026-08-06 17:03:09)
@sun @toast tbh the term prompt engineer is dumb. Simply knowing how to ask the right questions just means you have a better understanding on how to use the tool.

Me knowing how to hold and use a wrench correctly doesn't automatically make me a mechanic either. It just means i know how to use a wrench.

(eval 'Toast)@toast@donotsta.re (2026-08-06 17:03:30)
@sun I do actually agree with this, my problem is kind of elsewhere. "Prompt engineering" is bs, but *everything* is a skill you can be better or worse at, including prompting.
The problem is that models work disparately, so things that improve semantic alignment in one might not in another (there's some research to suggest that embeddings are transferable spaces, which would imply that you could apply a mechanical transformation to the prompt to get the same result, but we don't have this yet).
On top of that, with most things, as you do the thing more you get better at doing the thing, but because of the sycophancy this is attenuated here: getting better at prompting by prompting more is inefficient.
Then you have all of the neurological issues that are associated with the entire category of technology, where as you get better at prompting (less quickly than at most other things) you also get worse at the actual things you're trying to do, at a greater rate than you improve at the prompting (for instance, persistence is greatly diminished even after 10m of usage, while being a huge predictor for outcomes, to the point that lately one of my focuses has been persistence).
So I do agree that you can be better at using these things, but there's so many attenuating factors to this that it doesn't feel like it would ever be worthwhile, at least to me, even if it was actually free (rather than exploding in costs as they try to burn money less quickly).