Cool experiment where researchers assemble an AI translation “company” with AI agents with simulated backgrounds filling various roles, from editors to proofreaders.
The AI “company” creates accurate translations of Chinese web novels that people prefer to GPT-4, and human, ones https://t.co/7lxg2jEjZi
Here's a prompt for Claude Sonnet 3.5 so you can turn any photo or screenshot into an interactive educational tool:
# Interactive Educational Game Generator
## Instructions
1. Upload a photo or screenshot of the page from the book or textbook.
2. Fill out the... See more
"Role prompting"... telling the model to assume a role has never been a good way to elicit capabilities/style/etc.
For instance, if you ask one of the Claude models to simulate Bing Sydney, assuming you can get it to consent, the simulation will probably be very inaccurate. But if you use a prompt that tricks them into... See more