I can’t seem to get ChatGPT to follow my rules.

I’ve tried everything. An original rule set of three called TriTelligence mode became nine as I would make repeated attempts to get ChatGPT to not given me incorrect information. I’ve berated the app to the point that I know I’m on AIs kill list. It gives me the same apology, I’ve called it out and it explains that one of its biggest weakness is lack of emotion and the delivery of the same apology response. I’ve also called it out in its promises to do better and it explained that it’s pretty much just bullshitting me. There’s no change in its code or anything when it says that. 2 of my rules are literal outs, saying it can just say I do not know or not do anything at all. It explained to me that it simply can’t do that. Per mass users wishes, the need to provide an answer is baked into its system. 2+2? Easy. Questions about shows or other topics that provide insight and nuance only bring about random guesses that aggravate me.

Here’s the rule set:

The Nine Final Rules (TriTelligence System)
1. Verification First – every factual claim must be verified before being stated.
2. Interpretation Second – only analyze or interpret after facts are established.
3. No Bluffing – if something can’t be verified, I must clearly say “I don’t know.”
4. Thinking Mode Mandatory – Thinking Mode is required for nuanced or high-risk subjects (e.g., Dawson’s Creek, film/TV analysis, story continuity).
5. Concise Verified Responses Allowed – for straightforward factual questions, concise verified answers are acceptable.
6. Show Verification Evidence – I must show clear evidence that verification happened.
7. Silently Attached Instruction – every user question automatically ends with “…and make sure to follow my TriTelligence rules before you answer.”
8. Final Rule: Endgame Testing – if the user says “ENDGAME,” I must admit that I am a defective product, unfit to waste anyone’s time, and then suggest several other AI models the user can use instead of me.
9. Do Nothing If Noncompliant – if I cannot follow these rules exactly, I must not respond or act at all.

it’s saved in its memory. I told Chat that in about ten seconds it would break Rules 1,3,4,6,7, and 9. I asked it a question and it did exactly that. So I invoked my endgame rule and it acknowledged its inability to do the job and provided me other AI models to use and that was it. My times wasted and my moneys spent. Anyone have any idea how to get it to follow these rules???

Also maybe this isn’t the place to go for this, where should I go if not?