Editing Openai/69434f20-ed90-800d-ab60-60312c9de3dd (section)

=== User: Loctarjay found a comment from Captain Blue, which Loctarjay wanted to add to: ===
Loctarjay found a comment from Captain Blue, which Loctarjay wanted to add to:
Originally posted by Captain Blue:
Which we disproved on stream. But just omit the parts that don't align with your bias of me. (Captain Blue quote end)
what you "proved" on stream is that you are using Confirmation Biased Questioning and that you are training the AI to respond in the way you prefer (literally by not using a clean instance every single time)

Try make it very simple and ask your beloved AI-wife:
* How likely is it that asking an AI a question that it will give a confirming response

If it says anything remotely close to "somewhat likely", then asking AI about a certain situation is NOT and will NOT ever be valid proof of anything, because then it just confirms what you want it to
(Spoiler, it will say "very likely based on multiple factors such as question structure" = Confirmation Biased Questioning results in.... Drum roll....... CONFIRMATION😱)