In the modern day AI usage is the new norm and it's making AI users less patient. A year ago, I didn't mind repeating myself, rebuilding context, reminding a model what we talked about yesterday. Starting workflows from scratch felt fine. Now it feels weirdly frustrating. Not because the models are worse. Because once an AI remembers even a little, your expectations shift immediately.
I noticed this while using @OpenGradient Chat. The shift wasn't about whether the answers got smarter. It was how fast I started expecting it to pick up where we left off. One forgotten detail suddenly annoyed me way more than a mediocre response would have.
Once an AI carries context across interactions, something changes. Users stop evaluating individual prompts and start evaluating consistency over weeks, months. Memory stops feeling like a feature and starts feeling like an expectation. And expectations are harder to meet consistently.
I keep wondering which matters more: an AI that gives brilliant answers occasionally, or one that quietly remembers enough that you stop thinking about memory altogether?
@OpenGradient $OPG #OPG
I noticed this while using @OpenGradient Chat. The shift wasn't about whether the answers got smarter. It was how fast I started expecting it to pick up where we left off. One forgotten detail suddenly annoyed me way more than a mediocre response would have.
Once an AI carries context across interactions, something changes. Users stop evaluating individual prompts and start evaluating consistency over weeks, months. Memory stops feeling like a feature and starts feeling like an expectation. And expectations are harder to meet consistently.
I keep wondering which matters more: an AI that gives brilliant answers occasionally, or one that quietly remembers enough that you stop thinking about memory altogether?
@OpenGradient $OPG #OPG