If the crunch happens, wouldn't you profit from setting it up like a company? I guess if AI Alignment ends up a crisis that humanity only deals with last minute, but then with full effort, the funding will be there.
Get most of the people in the same office (relocate if necessary), have a company structure that is proven to work (hierarchies where necessary but not too much), get momentum going so you motivate each other to work longer and more productively.
Another consideration is that a lot of people that weren't doing AI Safety but were operating in a similar space might join the effort if the crunch happens, like Biontech developing the Covid vaccine although they were working on cancer before that.
This includes all big tech companies, so in case of a crunch the space of people doing Alignment Work might look very different.
Agree. I think most cases in modern worklife and productivity questions boil down to the mental fatigue + mental slowness components, which in my mind would be correlated but distinct.
Love the article!
I have ME/CFS so I have become hyper-aware of energy and fatigue, and how much it influences life outcomes in the small and big. For clinical cases it is obvious, but I think (and this is a bit speculative) that the "energy" aspect of it is underrated even in the healthy population. I think of it as "mostly upstream" of the pain, motivation, and classical self-control components. If you have energy, effort often doesn't feel unpleasant at all, so you wouldn't need to resist anything. Although I guess motivation and pain influence energy, not sure about the exact values. This would make classical self-control most downstream and often not the best point to change something.
Wondering about the science of it, haven't found a lot of good papers that look at "energy" in the non-clinical case (didn't even know what the correct domain term is). Could be a hard limit (like using up neuromodulators acetylcholine and epinephrine, and producing them faster would mess with the system in another way), which would favor genetic explanations of variation, or could be guided by beliefs and experiences.
I'm also wondering if the ME/CFS experience makes me see a less pronounced effect more clearly, or if the consequences in the interplay of energy-motivation-pain are shifted very much and it distorts my view on the non-clinical case.
Matches my experience with ME/CFS.
Another thing I like to do is look at things from a "comparative advantage" point of view:
I hope that I will get healthy at some point in my life, and I don't think that is unrealistic. Other than investing time into making that happen faster, which is only possible up to a certain degree (it might be the case that nothing works for me until there is a new drug in 10 years which then works, in which case only investing into health right now would be a mistake), I would like to do stuff that I can still somewhat do. So for personal development, invest precious energy in the thing you can still do at 25% and that's somewhat relevant for your life rather than the thing you can do at 5%.
Was the difference in mocking working on oneself and others statistically significant in the second survey? Effect looks somewhat small unless for ~60 participants.
Nice to see there is more variance in mocking working on oneself, which suggests there are different types more or less inclined to react to mockery. Also an intuitive result.
Thanks for doing the survey, I always appreciate when statements are backed up empirically 👍
Thank you for writing this, I had some of the same misconceptions.
A question about point 1, divestment: Does a similar principle apply when companies pledge to only use renewable energy? If they also build capacities so more renewable energy is generated, then obviously yes.
But otherwise I don't see how there is risk aversion in using more non-renewable energy for the other energy consumers that would shift the price.
Thanks for writing this.
By the way, I was thinking about how to find information on the internet for quite some time (the first Google results weren't that promising) and the Gwern post you linked looks exactly like the kind of thing I was looking for.