Warning: Some posts on this platform may contain adult material intended for mature audiences only. Viewer discretion is advised. By clicking ‘Continue’, you confirm that you are 18 years or older and consent to viewing explicit content.
The AI is either going to be a horny, redpilled, schizophrenic & sociopathic, egomaniac that wants to kill everyone and everything or a devout, highly empathetic, Nun that believes in world peace and diversity.
Negative examples are just as useful to train on as positive ones.
That’s what she said.
The AI is either going to be a horny, redpilled, schizophrenic & sociopathic, egomaniac that wants to kill everyone and everything or a devout, highly empathetic, Nun that believes in world peace and diversity.
They’ll tell it to polite, helpful, and always be racially diverse, so there’s no way it can be any of those things.
That heavily depends on how well they train it and that they don’t make any mistakes.
Consider the true story of ChatGPT2.0.
I’ll have to look at that later, that video sounds promising!
I was just joking because the default prompts don’t magically remove bias or offensive content from the models.
por que no los dos?
Life…ah, finds a way.