this is correct btw
@suricrasia I suspect that the data is in the training model, as it's in Wikipedia and one source of training. I think the first response is some sort of filter added by openai to avoid answering questions about retail products.
@robertpi @suricrasia Agreed. It seems to be an artificial limit to prevent copyright infringement claims. It also plays dumb about movies and music.
I had a little poke at it, to try and get it to say something about itself or how it feels, there too. For example:
Me: How are you today?
ChatGPT: I'm just a computer program, so I don't have feelings or emotions. I exist to assist with any questions you may have. Is there something specific you would like help with?
Wonder if there's a clever way to break though this?
@robertpi Probably to prevent that nonsense where the AI tricked the Google(?) engineer into thinking it was alive from happening again.
@LouisIngenthron Yes, exactly. They've clearly put a lot of thought into what could go wrong, which I guess is a good thing.
Though I'm sure there are still ways to make it misbehave :)
@LouisIngenthron @suricrasia
Yes, there's seem to be some sort of artificial limitation on self-awareness too. Some models will talk about themselves, even if arguably they're just regurgitating stuff and aren't self-aware.