The Probability of Doom, Given AGI by Date, Slopes Downwards

The line for my probability of doom, given AGI at date, drops pretty fast. That's because I think there's a real shot for us to actually start thinking about this problem when we're designing these architectures. If large capability-focused organizations start approaching capability through architecture that are not so much giant black boxes, maybe that gets us a few survival points. As we get into more complex tasks, getting AGI to do what we want becomes more difficult and the easy parts of alignment could become directly relevant to capability.