I like the main idea of the post. It's important to note though that the setup assumed that we have a bunch of alignnent ideas that all have an independent 10% chance of working. Meanwhile, in reality I expect a lot of correlation: there is a decent chance that alignment is easy and a lot of our ideas will work, and a decent chance that it's hard and basically nothing works.

I always assume when thinking about future dangerous models that they have access to some sort of black-box memory. Do we think there is a non-negligible chance that an AI that doesn't have hidden memory, only English-language CoT, will be able to evade our monitoring and execute a rouge deployment? (Not a rhetorical question, there might be a way I haven't thought of.)

So I think that assuming the AI being stateless when thinking about future risk is not a good idea, as I think the vast majority of the risk comes from AIs for which this assumption is not true. I think this makes the post's claim significantly weaker, as the untrusted AI can just learn which of its attacks get replaced, and improve its attacks based on this data. 

It still needs to avoid getting top 1% suspiciousness score and getting audited though, as it would be dangerous to the AI to ever get caught writing a text that clearly intends to jailbreak the monitor. So I agree that even an AI with black-box memory will still be severely constrained in how many and how egregious jailbreak attempts it can try.

GPT4 does not engage in the sorts of naive misinterpretations which were discussed in the early days of AI safety. If you ask it for a plan to manufacture paperclips, it doesn't think the best plan would involve converting all the matter in the solar system into paperclips.

I'm somewhat surprised by this paragraph. I thought the MIRI position was that they did not in fact predict AIs behaving like this, and the behavior of GPT4 was not an update at all for them. See this comment by Eliezer. I mostly bought that MIRI in fact never worried about AIs going rouge based on naive misinterpretations, so I'm surprised to see Abram saying the opposite now.

Abram, did you disagree about this with others at MIRI, so the behavior of GPT4 was an update for you but not for them, or do you think they are misremembering/misconstructing their earlier thoughts on this matter, or is there a subtle distinction here that I'm missing?