Episode
Confidential ML
Seth and I discussed confidential computing for AI, including how hardware trusted execution environments work, and how we can surface those guarantees for both inference and training.
Learn More:
The AI Show's Favorite links:
Seth and I discussed confidential computing for AI, including how hardware trusted execution environments work, and how we can surface those guarantees for both inference and training.
Learn More:
The AI Show's Favorite links:
Video URL
HTML iframe
Have feedback? Submit an issue here.