zlacker

While I do think this is fun and interesting, I agree that it is going to burn through a ton of tokens and there's a very serious risk that non-technical, and even some technical people probably, will interpret this as more than it is.

This is exactly the kind of thing llms are good at doing. Generating text based on other text.

And in this case, it's exactly the kind of text these models were trained on, so of course they're going to be very convincing.

Still, it is interesting to see where this technology is leading us and what fun side projects pop up along the way.