This looks like a really interesting session. The "inference economics" angle doesn't get talked about enough — a lot of AI ideas look great in demos but feel very different once cost, latency, and reliability enter the picture. Looking forward to hearing the real stories from teams building in production.
Out of curiosity, have you seen teams completely rethink a feature or product direction because of inference constraints?
Navid Kalaei supported this idea ·
This looks like a really interesting session. The "inference economics" angle doesn't get talked about enough — a lot of AI ideas look great in demos but feel very different once cost, latency, and reliability enter the picture. Looking forward to hearing the real stories from teams building in production.
Out of curiosity, have you seen teams completely rethink a feature or product direction because of inference constraints?