Someone wrote in [community profile] fandomsecrets 2023-11-06 12:36 am (UTC)

They are trained on a shitload of data scraped from the internet, but it's not done in real time. (Most likely, the data wasn't collected by hand by the company that created the model either.)

Post a comment in response:

If you don't have an account you can create one now.
HTML doesn't work in the subject.
More info about formatting