You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Apollo is a bit of a strange beast, with a mix of production and research code and mixed standards of code quality.
But as part of our Responsible AI policy, we need to ensure that any production AI is clearly and transparently documented.
Imagine a curious user coming over from Lightning to understand how the AI assistant works. It's really important that they can get a good high level overview of the codebase and get a basic understanding.
Stuff like which models are used, and when, and how, and with what prompts, and what data we use (if any) needs to be clearly explained. And not just for AI experts.
The text was updated successfully, but these errors were encountered:
Apollo is a bit of a strange beast, with a mix of production and research code and mixed standards of code quality.
But as part of our Responsible AI policy, we need to ensure that any production AI is clearly and transparently documented.
Imagine a curious user coming over from Lightning to understand how the AI assistant works. It's really important that they can get a good high level overview of the codebase and get a basic understanding.
Stuff like which models are used, and when, and how, and with what prompts, and what data we use (if any) needs to be clearly explained. And not just for AI experts.
The text was updated successfully, but these errors were encountered: