You wrote “AI has access to a wealth of our personal information from past conversations” - From a technical standpoint, AI has a “context” per session and it can have “memories” which can be added if the system you are using provides such a thing. However: Unless you are reusing the same session over several days, in general AI can’t remember your past conversations. And it’s not advisable to use the same session for an extended period of time: The context becomes diluted and the model is more prone to hallucinate, forget key points or start to misbehave in subtle ways.

On the content of your post… I’d wish you would be clearer and more precise in getting your point across. It was a hard read for me tbh. Also I think that “bypassing the brainstorming process” with AI is simply impossible, because a large language model does not have ideas. It’s output is the statistical mean of its training data, which in essence means that its output is by design mediocre. It’s the statistical middle ground of the options the training data contains. Human ideas on the other hand, as you rightfully point out are the result of a spark of personality and individual experiences. This will never be replaced by AI. Using it as a tool for brainstorming can be detrimental to the result for the same reasons. The AI might convince you that your idea is good while it is in fact just mediocre or worse.

I think you might unknowingly still be a victim of the mystification of LLMs. It’s time to see the technology at face value and to get to know its very real limitations of practical usability. The part about transhumanism… idk. Only people who see these things as more than just sophisticated word generators would entertain such an idea.