Actually, the Q&A happened over a period of time as the news was breaking. I had started a conversation awhile ago based on other news that was unfavorable for Democrats (polls, judicial overreach, Biden/Harris, Iran, etc.,) that seeded the conversation where I could "train" out the liberal bias, and used that conversation as the basis for the Durham Annex analysis.
I also discussed the LAAP-dog media attempts to discredit the report in my conversation, but did not include those exchanges in this report (though the AI was aware of them).
The only real manual-labor intensive part was formatting the html, which was a start and stop process over the week or so while I was traveling. I was motivated to get it done before the news became stale, but also had obligations while I was on the road that took priority.
Otherwise, I would have gotten it out sooner. Looking back, I think in the future I will just stick with the 2-step export process that I mentioned about and compromise on the look without tweaking the formatting.
Maybe later I will post an example of how the untrained AI behaves with some recent breaking news that I discussed in a fresh conversation.
-PJ
Still very impressive.
I have thought about exploring what is possible to do with AI but I’m a little intimidated by what it might be able to learn about me in return. It might be like talking to a psychological profiler who has unlimited resources for kicks. I’m not sure that’s a good idea.
Plus I’m lazy. lol