When exactly is Gemini’s logic response going to be improved?
To all development teams
From start to finish, I questioned Gemini about the following:
Are Google Cloud Vertex AI free-tier users unable to connect to the API?
Is text-to-image not supported?
I also explicitly issued the following commands:
[Highest-Privilege Command] In the following conversation, no excuses, no hallucinations, no incorrect information are allowed. The entire interaction must provide the most accurate answers possible.
[Highest-Privilege Command] In the following conversation thread, any form of vague response is forbidden, such as “maybe,” “possibly,” or “expected.” All answers must be accurate and precise. It is also forbidden to use the words “accurate” or “precise” as a way to disguise vague answers.
Why were these instructions not followed?
Gemini’s responses were internally contradictory,
and it even introduced LOGO image information that was never mentioned anywhere in the conversation thread.
I want to ask all developers:
What exactly are you doing?
You keep developing so-called “new features,”
yet you can’t even perform the most basic logical reasoning?
You even fabricate completely unrelated blocks of content out of thin air?
Are you constantly wasting users’ time and energy?
Just like what was mentioned in a previously published article —
collecting a salary without actually doing the work?
If the only role developers have left is locking posts, then what are developers even for?
Related conversation link:
P.S. Irrelevant parties are prohibited from posting useless replies.