Loading video…
Hook

Their other posts in the index, biggest breakout first.
That “it’s not about X, it’s about Y” line isn’t built into ChatGPT. It’s learned from us. When people think the AI misunderstood them, they often respond with “It’s not about this, it’s about that” to correct it. Over time, enough users repeat that structure that it becomes part of the patterns the model leans on when reframing an answer. Because the model is trained to mimic patterns from its training and interaction history, it now uses that phrasing as a default for emphasis or clarification. GBT users are just breaking GBT over and over again and poor Open AI just has to figure out how to fix it.