I think the lowest hanging fruit is 'don't repeatedly post publicly about how conservatives are odious people that we don't want to be even vaguely associated with'.

You might also enjoy this longer piece I shared here.

Or maybe you think that abortion bans seem 4 orders of magnitude more tractable than factory farming bans, which seems extremely unlikely to me. 

You might be interested in this excellent post by Ariel Simnegar, which argues that mandating fetal anesthesia for late-term abortions could be an effective and tractable intervention.

Thanks for sharing this detailed report, and most important for your important work keeping a potentially viable anti-pandemic technology legal!

I realized that if you were even arguing about abortion, then you must value human fetuses(which look a lot like chicken fetuses) 8,650 times more than tortured, murdered chickens.

This seems not at all true to me? Quite apart from my being skeptical about your maths, people are allowed to care and argue about things that aren't as important as factory farming. Very few people spend all their effort on the single most important cause. To be honest, this seems like an isolated demand for rigour.

I think that centralisation (by which I assume you really mean OP-funding-centralisation) is a contingent fact about the EA movement, rather than an inherent one. And it sounds like you agree. But then I'm not sure why we'd use this as an exclusion criteria? If nothing else, if, once centralised, no a group being quite independent is sufficient alone for exclusion, then you can basically never decentralise.

Oh wow, I actually think your grandparent comment here was way more misleading than their tweet was! It sounds like they almost verbatim quoted you. Yes, they took out that you set up the experiment... but of course? If write "John attempted to kill Sally when he was drunk and angry", and you summarise it was "John attempted to kill Sally, he's dangerous, be careful!" this is a totally fair summarisation. Yes it cuts context but that is always the case - any short summarisation does this. 

In contrast, unlike your comment, they never said 'escape into the wild'. When I read your comment I assumed they had said this.

Also, the tweet direct quotes your tweet, so users can easily look at the original source. In contrast your comment here doesn't link to their tweet - before you linked to it I assumed they had done something significantly worse.

in response to our recent paper "Alignment Faking in Large Langauge Models", they posted a tweet which implied that we caught the model trying to escape in the wild. I tried to correct possible misunderstandings here.

Probably would be easier for people to evaluate this if you included a link?

Thanks for the comment! You're right that this approach would need modification if 'dangers that only become apparent after mass deployment' becomes a major risk factor, and that a 'trial' commercialisation period could be a good response. My hope is that the regulatory exam period would be able to catch much more than at present though - the regulator would have ample time to design and deploy more sophisticated tests, with the aid of labs who would presumably love to submit a test their competitor would fail (so long as they themselves pass).

If EA was a broad and decentralised movement, similar to e.g., environmentalism, I'd classify SMA as an EA project. But right now EA isn't quite that. Personally, I hope we one day get there.  

This seems pretty circular to me?

Interesting suggestion! Continuous or pseudo-continuous threshold raising isn't something I considered. Here are some quick thoughts:

• Continuous scaling could make eval validity easier, because the jump between eval-train (n-1) and eval-deploy (n) is smaller.
• Continuous scaling encourages training to be done quickly, because you want to get your model launched before it is outdated.
• Continuous scaling means you give up on the idea of models being evaluated side-by-side.