LLMs are bullshitters. But that doesn't mean they're not useful
1–10 of 59 posts
Re: LLMs are bullshitters. But that doesn't mean they're not useful
#2The problem is we can't label them as such. If they're bullshitters, then let's call it a LLBSer. It has a nice ring to it. Good luck with your government funding asking for another billion for a bullshitting machine bailout.
Re: LLMs are bullshitters. But that doesn't mean they're not useful
#3The problem is we can't label them as such. If they're bullshitters, then let's call it a LLBSer. It has a nice ring to it. Good luck with your government funding asking for another billion for a bullshitting machine bailout.
"BS in Computer Science" hits different
Re: LLMs are bullshitters. But that doesn't mean they're not useful
#4Good article, I just shared it with my non-technical family because more people need to understand exactly this about AI.
Re: LLMs are bullshitters. But that doesn't mean they're not useful
#5> You should not go to an LLM for emotional conversations
I'm more worried about who's keeping track of what's being shared with LLM's. Even if you could trust the model to respond with something meaningful, it's worth being very careful how much of your inner thoughts you share directly with a model that knows exactly who you are.
Re: LLMs are bullshitters. But that doesn't mean they're not useful
#6Same goes for many people.
Re: LLMs are bullshitters. But that doesn't mean they're not useful
#7Every time people post these 'gotcha' LLM failures, they never work when I try them myself.
E.g. ChatGPT has no problem with the surgeon being a dog: https://chatgpt.com/share/691e04cc-5b30-800c-8687-389756f36d...
Neither does Gemini: https://gemini.google.com/share/6c2d08b2ca1a
Re: LLMs are bullshitters. But that doesn't mean they're not useful
#8I've come to cease all "inquiry" type usage of LLMs because of this. You really can't trust anything they say at all that isn't verified by a domain expert. But I can let it write code for me, and the proof is in the PR. I think ultimately the real value in these things is agentic usage, not knowledge generation.
Re: LLMs are bullshitters. But that doesn't mean they're not useful
#9Every time people post these 'gotcha' LLM failures, they never work when I try them myself. E.g. ChatGPT has no problem with the surgeon being a dog: https://chatgpt.com/share/691e04cc-5b30-800c-8687-389756f36d... Neither does Gemini: https://gemini.google.com/share/6c2d08b2ca1a
This is like the LLM era version of the search bubble that prevented people from having the same search results for ostensibly identical searches.
Also keep in mind that LLMs are stochastic by design. If you haven't seen it, Karpathy's excellent "deep dive into LLMs like chatgpt" video[0] explains and demonstrates this aspect pretty well:
Re: LLMs are bullshitters. But that doesn't mean they're not useful
#10The headline feels like a strawman.
LLMs are very useful. They are just not reliable. And they can't be held accountable. Being unreliable and unaccountable makes them a poor substitute for people.