this post was submitted on 03 Jun 2025
382 points (97.8% liked)
Technology
70916 readers
3567 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
I've been trying to configure ChatGPT tell me if I'm wrong in a question or statement but damn it never does unless I keep probing for support or links. I've been having the feeling that it has become worse with latter models. Glad but also sad to see I was right.
Anybody know other LLM that are more "trustworthy"* and capable of searching online for more information?
Edit; *trustworthy in quotes because of course people will jump on this. I know the limitations of LLM, I don't need you to tell me how much you hate everything AI. And I know LLM aren't AI.
There are no trustworthy LLMs. They don't know it understand what they're saying - they're literally just predicting words that sound like they match what it was taught. It's a only barely smarter than a parrot, and it has no idea how to research anything or tell facts from made-up bullshit. You're wasting your time by trying to force it to do something it's literally incapable of doing.
You're better off researching them the hard way; check primary sources and then check the credibility of those sources.
Considering that parrots can have actual thoughts, I'd say LLMs are even less smart than that.
Claude definitely has its impressive moments where it calls out something inaccurate.
It’s also way less sycophantic, mature and better for light coding.
My only issue is that the servers are sometimes slow and so is the ios app which frequently trows an error after 2 minutes if waiting.
Claude 3.7 told me i'm wrong a couple of times. It knows how to search. I don't have an opinion on 4 yet but it can search too
On my Lemmy?? Impossible. I was just told by three different people rather condescendingly and assuming just as much about me that I was suffering from AI brain rot for daring to compare something to AI. The horror.