I have learned that AI safety is a big issue for 5 years now. Why do I only do anything now? There are a bunch of factors, but I think one really important factor was me finding other people making mistakes. Because this happened recently, the version of me before...
I am writing this using a dictation tool. The dictation tool will make mistakes, and there is a good chance that I will post it before correcting the mistakes. Picture me just walking and speaking, and the lightly edited transcription of that is what you're reading. I think I'll burst...
With optimisation-robust I mean that it withstands point 27 from AGI Ruin: > When you explicitly optimize against a detector of unaligned thoughts, you're partially optimizing for more aligned thoughts, and partially optimizing for unaligned thoughts that are harder to detect. Optimizing against an interpreted thought optimizes against interpretability. Are...
Public tripwires are no tripwires. I'm writing a quick and dirty post because the alternative is that I wait for months and maybe not write it after all. I am broadly familiar with the state of interpretability research but do not know what the state of model evaluations is at...
My AI pessimism is not addressed by optimistic arguments that do not promise the level of reliability required for a technology that our societies vital functions are expected to rely on. Historically, useful improvements are hard to reverse and solutions can suddenly fail after we had become dependent on them....