The OP is about Chicago.

if you generate a lot of output text you can approximate the hidden state.

You have to input the title of the article you want, not instructions like "write me ..."

I did. Eventually I managed to get an article, but by then it was generic enough to not be particularly useful.

Which is even more interesting since that could pointing to older unknown works.

I saw nothing indicating that; is there a source?

It's September.

How large is the model in terms of parameter numbers? There seems to be zero information on the size of the model.

It's not about the raw number of people, it's more about how old and aging they are. Age distribution is the key factor. 1 billion 60 year olds is not the same as 1 billion 20 year olds.

ArXiv link: https://arxiv.org/abs/2310.02207

You seem to be looking for a topic model. BERTopic might help: https://maartengr.github.io/BERTopic/index.html#quick-start

This is great...thank you!