So ChatGPT can discover facts about itself and add them to the context. Neat. There is now a lot of writing about ChatGPT on the internet. What happens when th

So ChatGPT can discover facts about itself and add them to the context. Neat. There is now a lot of writing about ChatGPT on the internet. What happens when that is included in the training set? I'd expect a similar outcome... it knows some facts about itself. Interesting to see how this would bias its behavior or the ability to run experiments like this one.