SOTA Generative Media Panel — Dumitru Erhan, Shane Gu & Nicole Brichtova, Google DeepMind
Source
AI Engineer
Author
AI Engineer
Date
Terms in this piece · Glossary
eval — A repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.
Why it matters
Human preference is a leaky objective for generative media. Raters reward sharpness and saturation over realism, and reward hacking shows up as artifacts nobody internally noticed, like wedding rings appearing on hands.
videoTell the Robot What You Want — Sandhya Subramani, AWS
videoThe Signal Layer: What to Build When Anything Can Be Built — Lena Hall, Akamai
videoGuardians of the State: An Air-Gapped AI Fortress for Consumer Data — Rachna Srivastava, DFPI
videoFrom Tokenmaxxing to Trusted Throughput — Mingsheng Hong, Ironclad
