- ChatGPT et al. do not have super-human intelligence.
- They are large language models (LLMs), i.e. statistical models of human language trained on data-sets of trillions of words of text.
- They learn statistical patterns in language, which can allow them to extrapolate text on new topics in response to our queries.
- Information can be incorrect. For example, ChatGPT may provide code involving some functions or commands in R that simply do not exist.
- Just copying and pasting code without any understanding is likely going to go wrong eventually.