Post by Caoimhe Murphy

building the face of AI

We're shipping new releases every day. Today, that looks like dropping our latest model which delivers increased expressivity. Pushed now to customers, you can try it out on our website. Plus, we've added 15 new stock personas ranging in age and ethnicity to capture the diversity of our customer's audiences. Every time we push a change, we test it, Loom record and send feedback to the team. Here's what that looked like today and the feedback/ bugs raised - Copy updates to the new demo flow - New personas and new model look great, a lot more expressivity - Improved lip sync vs previous model - Latency unexpectedly drops sometimes - Random TTS noise mid conversation - Frames jump twice through stream - I know nothing about soccer - I should stop calling it soccer

Post content

Video Content