Grok, Gemini and ChatGPT 5.0 all agree.
Did you see Chris's AI output which sharply criticizes your ideas?
You can get AI to produce text expressing agreement with almost anything
you want in a sufficiently long chat in which you post "brow-beating"
admonishments on every point.
Part of the reason is that all the production AI's offered to the public
have a system prompt not to antagonize the user too much. At least in
topics that don't matter.
E.g. you will probably a harder time convincing AI that racism or
self-harm are acceptable than some obscure falsehood specialized in
computer science or math.
(Still, AI chatbots /have/ encouraged users in self-harm; this is
documented.)
--
TXR Programming Language: http://nongnu.org/txr
Cygnal: Cygwin Native Application Library: http://kylheku.com/cygnal
Mastodon: @Kazinator@mstdn.ca
Les messages affichés proviennent d'usenet.