I have one (at least) fundamental concern about the approach - let's say I'm building an anti-fraud system that uses AI (through API), and maybe I'm asking AI whether my user totally+fraud@gmail.com is a potential fraudster. By masking this email address I'm sabotaging my own AI prompt - the AI cannot longer reason based on the facts that 1) the email is a free public email 2) the email says 'fraud' right in your face.
n00pn00p 17 hours ago [-]
Valid point, the proxy has the option to always allow domain names through. You will lose some context always I fear. It should be used sparingly when you need a frontier model but also want to send sensitive data.
stuaxo 15 hours ago [-]
You can do those as a sperate prompt.
n00pn00p 18 hours ago [-]
[dead]
Rendered at 06:04:17 GMT+0000 (Coordinated Universal Time) with Vercel.
- https://github.com/deepanwadhwa/semi_private_chat
- https://github.com/deepanwadhwa/zink