Remix.run Logo
smlacy 8 months ago

I find the "Can you ..." phrasing used in this demo/project fascinating. I would have expected the LLM to basically say "Yes I can, would you like me to do it?" to most of these questions, rather than directly and immediately executing the action.

jasonjmcghee 8 months ago | parent | next [-]

If an employer were to ask an employee, "can you write up this report and send it to me" and they said, "yes I can, would you like me to do it?", I think it would be received poorly. I believe this is a close approximation of the relationship people tend to have with chatgpt.

pbhjpbhj 8 months ago | parent [-]

Depends, the 'can you' (or 'can I get') phrasing appears to be a USA English thing.

Managers often expect subordinates to just know what they mean, but checking instructions and requirements is usually essential and imo is a mark of a good worker.

"Can you dispose of our latest product in a landfill"...

Generally in UK, unless the person is a major consumer of USA media, "can you" is an enquiry as to capability or whether an action is within the rules.

IME. YMMV.

iamjackg 8 months ago | parent | prev | next [-]

I'm very curious why you think that! Sincerely. These models undergo significant human-aided training where people express a preference for certain behaviours, and that is fed back into the training process: I feel like the behaviour you mention would probably be trained out pretty quickly since most people would find it unhelpful, but I'm really just guessing.

lolinder 8 months ago | parent | prev [-]

What distinguishes LLMs from classical computing is that they're very much not pedantic. Because the model is predicting what human text would follow a given piece of content, you can generally expect them to react approximately the way that a human would in writing.

In this example, if a human responded that way I would assume they were either being passive aggressive or were autistic or spoke English as a second language. A neurotypical native speaker acting in good faith would invariably interpret the question as a request, not a question.

pbhjpbhj 8 months ago | parent [-]

In your locality.

I've asked LLM systems "can you..." questions. I'm asking surely about their capability and allowed parameters of operation.

Apparently you think that means I'm brain damaged?

gnabgib 8 months ago | parent | next [-]

Surely there's better Windmills for you to tilt at.

pbhjpbhj 7 months ago | parent [-]

For sure.

It's basically an observation on expectations wrt regional language differences. HAND.

int_19h 7 months ago | parent | prev [-]

LLMs are usually not aware of their true capabilities, so the answers you get back have a high probability of being hallucinated.

pbhjpbhj 7 months ago | parent [-]

So far, they seem to be correct answers.

I assume it's more a part of explicitly programmed set of responses than it is a standard inference. But you're right that I should be cautious.

ChatGPT, for example, says it can retrieve URL contents (for RAG). When it does an inference it then shows a message indicating the retrieval is happening. In my very limited testing it has responded appropriately. Eg it can talk about what's on HN front page right now.

Similarly Claude.ai says it can't do such retrieval - except through API use? - and doesn't appear to do so either.