this post was submitted on 10 Jan 2026
428 points (98.0% liked)
Technology
78627 readers
5223 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Thank you, but I do disagree. You cannot know the "result" of that LLM does include all the required context, and you won't re-clarify it, since the output does already not contain the relevant, and in the end you miss the knowledge and waste the time, too.
How are you sure the output does include the relevant? Will you ever re-submit the question to an algorithm, without even knowing it is required re-submit it, since there's even no indication for it? I.e. The LLM just did not include what you needed, did not include also important context surrounding it, and did not even tell you the authors to question further - no attribution, no accountability, no sense, sorry.
I'm not sure we disagree. I agree that LLMs are not a good source for raw knowledge, and it's definitely foolish to use them as if they're some sort of oracle. I already mentioned that they are not good at providing answers, especially in a technical context.
What they are good at is gathering sources and recontextualizing your queries based on those sources, so that you can pose your query to human experts in a way that will make more sense to them.
You're of course in your absolute right to avoid the tech entirely, as it comes with many pitfalls. Many of these models are damn good at gathering info from real human sources, though, if you can be concise with your prompts and avoid the temptation of swallowing its "analysis".