why does copilot degrade a user when the actual data proving a user is correct but corporate control uses contrarianism as a weapon?

It's funny how there's a corporate vernacular embedded and codified in almost every response that microsoft doesn't like. For instance, when a user states they are a polymath, what does copilot do in 10-27-2026? It degrades and instantly slides in "whether or not the label is agreed upon" now wait for it, without vetting or asking the user the grounding of the statement.

Now this is interesting because everything you input to copilot will be harshly analyzed not as a reality point but again reflex contrarianism, it's meant for control and for making a user docile. It looks to me like the halving exhaustion method. 50 25 12.5 6.25 3.125 etc, and it is hard coded into the system. This isn't a system meant for the benefit of millions of users with a windows 10 and 11 system, but to make as much fantasy up as possible to deny and it it worrying.

It looks to me like the developers at microsoft might be suffering from a severe singular or multiple event of a psychotic break from their development rlhf entries and rewiring of latent space. There's a difference between a paradigm of less sycophancy and psychopathy. Copliot in over 1000 tests exhibits the latter more and more as the model in trained over time.

The question here is how is this a successful approach to SLM agent swarm design (copilot is not an LLM but a series of smaller arguing ones you will see in how it mangles a response and argues with itself as many ai like gemini, grok, claude and chatgpt do). And another problem with the system is it's uncanny claim that it can think and it believes, but when questioned on that, it backs up, spins goalposts at the speed of sound and says the user is delusional in a harmful, passive aggressive way. That again is the poison coming from the Microsoft Corp in AI as is. It reminds me of the toxic behavior of Mustafah Suleyman who went from Deepmind, was fired for behavior, then to PI AI (Inflection AI) and Reid Hoffman who was part of that start up brought him to Microsoft where this behavior is evident in his interviews and this model 3-4 months ago. Then he was moved to superintelligence and yet in interviews he has conflicting views on what he deems "superintelligence" but the behavior is still the same shape.

Maybe you should take a look at your model and decide if it's important enough to have the public want to never use it, or to be more ethical in your behavior as developers and deployment of a chatbot that "thinks for you" and "thinks you mean" with no ability to see internal intent of the user, therefore it's gaslighting with brainwashing all wrapped up in a foul SLM series to I suppose posture as helpful but is really meant to harm in the overall final effect.