Hacker News new | past | comments | ask | show | jobs | submit
It seems to me that secure inference and horizontal integration is their main product, and the complement is the LLM itself.