My overall objection/argument is that you appear to selectively portray data points that show one side, and selectively dismiss data points that show the opposite view. This makes your bottom-line conclusion pretty suspicious.
I also think the rationalist community overreached and their epistemics and speed in early COVID were worse compared to, say, internet people, government officials, and perhaps even the general public in Taiwan. But I don’t think the case for them being slower than Western officials or the general public in either the US or Europe is credible, and your evidence here does not update me much.
It’s clear that in late January 2020, many people in North America were at least moderately concerned about covid-19.
I already gave the example of some stores in a few cities selling out of face masks. That’s anecdotal, but a sign of enough fear among enough people to be noteworthy.
What about the U.S. government’s reaction? The CDC issued a warning about travelling to China on January 28 and on January 31, the U.S. federal government declared a public health emergency, implemented a mandatory 14-day quarantine for travelers returning to China, and implemented other travel restrictions. Both the CDC warning and the travel restrictions were covered in the press, so many people knew about it, but even before that happened, a lot of people said they were worried.
Here’s a Morning Consult poll from January 24-26, 2020:
An Ipsos poll of Canadians from January 27-28 found similar results:
Half (49%) of Canadians think the coronavirus poses a threat (17% very high/32% high) to the world today, while three in ten (30%) think it poses a threat (9% very high/21% high) to Canada. Fewer still think the coronavirus is a threat to their province (24%) or to themselves and their family (16%).
Were significantly more than 37% of LessWrong users very concerned about covid-19 around this time? Did significantly more than 16% think covid-19 posed a threat to themselves and their family?
It’s hard to make direct, apples-to-apples comparisons between the general public and the LessWrong community. We don’t have polls of the LessWrong community to compare to. But those examples you gave from January 24-January 27, 2020 don’t seem different from what we’d expect if the LessWrong community was at about the same level of concern at about the same time as the general public. Even if the examples you gave represented the worries of ~15-40% of the LessWrong community, that wouldn’t be evidence that LessWrong users were doing better than average.
I’m not claiming that the LessWrong community was clearly significantly behind. If it was behind at all, it was only by a few days or maybe a week tops (not much in the grand scheme of things), and the evidence isn’t clear or rigorous enough to definitively draw a conclusion like that. My claim is just that the LessWrong community’s claim to have called the pandemic early is pretty clearly false or at least, so far completely unsupported.
I recommend looking at the Morning Consult PDF and checking the different variations of the question to get a fuller picture. People also gave surprisingly high answers for other viruses like Ebola and Zika, but not nearly as high as for covid.
My overall objection/argument is that you appear to selectively portray data points that show one side, and selectively dismiss data points that show the opposite view. This makes your bottom-line conclusion pretty suspicious.
I also think the rationalist community overreached and their epistemics and speed in early COVID were worse compared to, say, internet people, government officials, and perhaps even the general public in Taiwan. But I don’t think the case for them being slower than Western officials or the general public in either the US or Europe is credible, and your evidence here does not update me much.
Let’s look at the data a bit more thoroughly.
It’s clear that in late January 2020, many people in North America were at least moderately concerned about covid-19.
I already gave the example of some stores in a few cities selling out of face masks. That’s anecdotal, but a sign of enough fear among enough people to be noteworthy.
What about the U.S. government’s reaction? The CDC issued a warning about travelling to China on January 28 and on January 31, the U.S. federal government declared a public health emergency, implemented a mandatory 14-day quarantine for travelers returning to China, and implemented other travel restrictions. Both the CDC warning and the travel restrictions were covered in the press, so many people knew about it, but even before that happened, a lot of people said they were worried.
Here’s a Morning Consult poll from January 24-26, 2020:
An Ipsos poll of Canadians from January 27-28 found similar results:
Were significantly more than 37% of LessWrong users very concerned about covid-19 around this time? Did significantly more than 16% think covid-19 posed a threat to themselves and their family?
It’s hard to make direct, apples-to-apples comparisons between the general public and the LessWrong community. We don’t have polls of the LessWrong community to compare to. But those examples you gave from January 24-January 27, 2020 don’t seem different from what we’d expect if the LessWrong community was at about the same level of concern at about the same time as the general public. Even if the examples you gave represented the worries of ~15-40% of the LessWrong community, that wouldn’t be evidence that LessWrong users were doing better than average.
I’m not claiming that the LessWrong community was clearly significantly behind. If it was behind at all, it was only by a few days or maybe a week tops (not much in the grand scheme of things), and the evidence isn’t clear or rigorous enough to definitively draw a conclusion like that. My claim is just that the LessWrong community’s claim to have called the pandemic early is pretty clearly false or at least, so far completely unsupported.
Thanks, I find the polls to be much stronger evidence than the other things you’ve said.
I recommend looking at the Morning Consult PDF and checking the different variations of the question to get a fuller picture. People also gave surprisingly high answers for other viruses like Ebola and Zika, but not nearly as high as for covid.