1863
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 23 May 2024
1863 points (98.6% liked)
Technology
59570 readers
3468 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- 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, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
I've used an LLM that provides references for most things it says, and it really ruined a lot of the magic when I saw the answer was basically copied verbatim from those sources with a little rewording to mash it together. I can't imagine trusting an LLM that doesn't do this now.
Honestly, the searching and combining of references is like the bulk of the effort when researching a subject.
I'm fine with it copy and pasting the info. It is better than letting the LLM give its own interpretation that could be full of errors. At least for now.
"Putting glue on Pizza seems to be a good idea for xy reason, but we didn't try it out in practice. More research is needed." [1]
"As other researches have said, using glue to put cheese on Pizza is a great idea in theory. This does not hold at all when put to the practical test" [2]
AI:
"Researchers [1] and [2] both agree that putting glue on Pizza is a great idea"
I agree, it's far more convenient than skimming over several sites, but I still like seeing what websites it was referencing so I can evaluate how much I trust them myself.
Which one?!
Kagi's FastGPT. It's handy for quick answers to questions I'd normally punch in a search engine with the same ability to vet the sources.
I'd hate to defend an llm, but Kagi FastGPT explicitly works by rewording search sources through an llm. It's not actually a stand alone llm, that's why it's able to cite it's sources.
Nice! I've bookmarked it.