4 ms·
How to use ChatGPT to fact-check itself
- sebjag 3y agointeresting
- al2o3cr 3y agotwo random prompts are somewhat unlikely to remain consistent in their answers unless there is some underlying correlation handwaving intensifies Seems like this also pushes the tricky decisioning into two other places: which pieces of the prompt to mask, and whether a second answer is "consistent"