• 13 Posts
  • 537 Comments
Joined 2 years ago
cake
Cake day: October 28th, 2023

help-circle





  • A large language model is trained on language that we have created. It is programmed to succeed a given task.

    It tries to create an output with the highest success score based on the numbered weights of certain combinations of words, versus the weights of the combinations of words given as an input. These weights are based on our language, our literature. Part of that literature includes crime.

    It doesn’t have realisations, or awareness, or a motivation, or a sense of self preservation; it’s an incredibly complex calculator that runs an input through an algorithm based on our language and then generates an output.

    The output score for blackmail and murder is probably very high based on the given input.