The Deep Mind Alignment Team's Threat Model

DeepMind has released a new AI model that it says could lead to catastrophic or existential risk from AI. It's the third in a series of three fairly significant announcements made by leading AI labs over the last few weeks on this topic. The idea is you may actually think that you're training an AI to do next word prediction or whatever beyond a certain threshold of capability it actually develops distinct goals and those goals are intrinsically potentially unpredictable so anyway kind of an interesting technical dive into how the mind is thinking about this or at least their lengthYeah i think it is quite interesting it is pretty short presentation but i think similar to anthropic it is very concrete about sort of the approach we're taking it,