Monday Links 13

Learning from the incident you didn’t have

When an incident happens in an organization, the traditional response is to identify ways to prevent the incident from happening again in the future. The community around this website takes a different approach towards incident analysis. To paraphrase the late computer scientist Edsger Dijkstra, incident analysis is no more about incidents than astronomy is about telescopes. Instead of focusing on prevention, we seek to leverage incidents as an opportunity to learn as much as possible about how work is done within the organization.

Import AI Newsletter on the discussion about the safety of releasing AI models without restriction

Rep. Anna Eshoo (a Democrat from California) has sent a letter to the White House National Security Advisor and Office of Science and Technology Policy saying she has “grave concerns about the recent unsafe release of the Stable Diffusion model by Stability AI”. The letter notes that Stable Diffusion can be used to generate egregiously violent and sexual imagery, and - due to eschewing the kinds of controls that OpenAI uses for its commercial product DALL-E2 - the freely accessible model represents a big problem.

Letters like this are indicative of a culture war brewing up among AI researchers; on one side, groups want to slowly and iteratively deploy new technologies via APIs with a bunch of controls applied to them, while on the other side there are people who’d rather take a more libertarian approach to AI development; make models and release the weights and ride the proverbial lightning.