Hey, Michael, I was actually curious to try this, if you actually have something written up that sort of "primes" the AI's interpretation of the world.

Do you have a link to the paper so I can try what you suggest?