A 2020 study in Psychology Research and Behavior Management found that effective impression managers - people who consistently project a favorable self - receive less support from friends and colleagues, because others take their competence at face value.
Now, a parallel pattern is emerging in AI interaction: people are telling chatbots things they won't say to anyone who knows them. The two findings connect directly. The audience that needs you to appear fine is also the audience you cannot be honest with.
In tech, the performance is relentless. You learn to answer before you fully process. Hesitation reads as weakness. Certainty becomes automatic. The capable version of you assembles before you decide to put it on.
The trap is not that people don't care. It's that a well-maintained image becomes its own barrier. People stop seeing through it - not because they aren't paying attention, but because you are very good at this.
Disclosure to AI is easier because there are no social consequences. No colleague files it away. No partner references it weeks later. No reaction that quietly reorganizes the space between you.
For those exhausted from performing a particular self, that absence is not incidental. It is the specific thing that makes honesty possible.
The relief is real. But if the easiest place to be honest is a context with no stakes and no memory, then the contexts that matter - the partner, the friend, the colleague - have become places where honesty is too expensive.
That is not a story about AI. It is a story about what professional life has done to the rest of life.