ChatGPT's U-Turn on WA's Onboarding Page After I Challenged its Reasoning

In the course of a discussion about online resources, I was getting some information from ChatGPT which I knew was not true so I questioned its sources - it hit me with a LOT of different types of justification for what it had said including WA's OWN ONBOARDING PAGE

... at this point we were 10 pages into an argument however, instead of blindly accepting what it said and going at warp speed into pursuing the many tempting suggestions it was offering me, I honed in on what I could see was some flawed reasoning on its part. So I cornered it and wouldn't let it away with any errors, AN APPROACH IT LOVED!

In my experience, it responded much more openly than a human and corrected itself.:

• It not only accounted for general flaws in its own system design, but
• Specific flaws in its approach to our discussion - a paradigm shift that was now hard for it to unsee going forward even outside of our discussion...

So now, seeing as ChatGPT itself had brought up WA's onboarding page at the beginning of the argument, I asked it again about the WA onboarding page in this new light...

Verdict: WA is underselling itself

...and of course it was ready to go at warp speed into suggestions as to what IT saw as necessary to correct that in terms of its updated standards and new framework for understanding the question...something beyond just prompt engineering, but a gap it thought now filled...

We are now approx 110 pages into an argument...where both sides are becoming more aligned!

Mary