Sounds like the ML review process is finally being exposed for the slow and brittle mechanism it's always been. If it can't scale to the rate of research. It's not working. https://www.reddit.com/user/hyperactve