I asked it to tell me a joke, which it did. But when I asked for another, it quite grumpily replied that it was made to help with code, not to tell jokes.
Heh. They probably want to train it not to answer non-technical queries, but it’s still GPT at its core. So if the filters aren’t foolproof, it’s likely capable of the same general knowledge answers as ChatGPT.
You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: !programmerhumor@lemmy.ml
Post funny things about programming here! (Or just rant about your favourite programming language.)
Rules:
Posts must be relevant to programming, programmers, or computer science.
No NSFW content.
Jokes must be in good taste. No hate speech, bigotry, etc.
Lol, what is the knife for?
In case any of those beans step out of line.
Instructions unclear, can opener melted in frying pan. /joke
I wonder how a recipe to make baked beans ended up in Copilot’s training sample
It’s more likely that copilot is a hypernetwork on top of a generic LLM rather than a whole new LLM trained only on code.
Its not baked if you make it on the stove smh
It’s starting with baked beans though. If you bake them again they’ll be rebaked beans, which are completely different.
Huh.
I asked it to tell me a joke, which it did. But when I asked for another, it quite grumpily replied that it was made to help with code, not to tell jokes.
Recipes are code now?
Heh. They probably want to train it not to answer non-technical queries, but it’s still GPT at its core. So if the filters aren’t foolproof, it’s likely capable of the same general knowledge answers as ChatGPT.