I think offsets are an excellent way to keep some cause-promoters honest. For instance, if people who care about animal welfare are tempted to exaggerate the effectiveness or evidence base for animal charities, they might be deterred by the thought that people will make the obvious inference about offsets, and conclude that it's not worth it to give up animal products because it's worth more to them than the price of the offset.

It seems equally plausible that otherwise honest cause-promoters would be incentivized to be dishonest and downplay their cause effectiveness. In general, I don't think that assuming that everyone is a rational economic actor and speculating on their incentives to lie is very productive.

This works even if no one actually buys the offset - you can use this kind of number to help establish a preference ordering among very different uses of resources, like Katja Grace does here.

It's also worth just running the thought experiment as a check on your numbers - would you actually be happy if people gave $10 to CiWF instead of giving up chicken for a year? At what scale does this change?

Sure, that's fine, just not offsetting.

I disagree with the idea that donating exactly $1 per day to one kind of charity and making other donations to other charities is the best allocation of resources you can make for any reasonable value function. If there is a better cause out there, then it deserves that one dollar too.

This is silly, if reducing animal suffering is a priority then you ought to do it rigorously; if it is not a priority then you ought to focus predominantly on something else.

I think it's good to have our high standards of evidence, but once in a while I see someone with a good idea that just gets shrugged off. For instance it seemed like Denkenberger's paper on alternative foods was mostly ignored, even though it gave a pretty good conclusion that his idea was better in expectation than poverty relief.

There just needs to be more updating. When people see that, they will be encouraged to be creative because they'll have more confidence that if they get a good idea then it will attract attention.

I don't think people are properly grasping what it would mean to have a set of shelters on Earth that would be equally well funded as a permanent self-sustaining colony off-Earth. You could probably afford equally-sized self-sustaining colonies in underground locations in each of multiple different climate zones as well as multiple undersea locations. Plus each of them could be better hardened in all sorts of ways. But it's nearly impossible to estimate because a permanent, self-sustaining extraterrestrial colony is something with almost unbounded size and complexity. We currently don't know how to synthesize all required materials from nearby planets and the number of people required could easily be tens of thousands.

To be honest, spending that much money on shelters on Earth is downright absurd. Instead of such a monstrous shelter-building program, you could spend 5% of the money on shelters which would be 95% as effective at increasing the probability that humanity would survive a catastrophe. That's the more likely and relevant comparison.

In principle it's very important, but I'm a little more skeptical about the value of actually doing it because I think it's very intractable.