This is a harmful post, and should be deleted or removed.
Was outside of LW norms. It came off as a blunt attempt to shut down discussion, with very little in terms of justification for doing so. This is in no way a clear cut infohazard, and even if it was, I’m not convinced that shutting down discussion of things that might be infohazards is a good policy, especially on a relatively obscure site centered around truth seeking. Statements this confident about issues this complicated should only be said after some extensive analysis and discussion of the situation. jtm’s presentation of the issue struck me as far more tempered and far less adversarial. I’d encourage Davidmanheim to supplant his comment with a more fleshed out version of his position.
I am shocked to hear that people need proof something is an infohazard before deciding that the issue needs to be discussed BEFORE posts like this go live. I see no evidence that any such discussion occurred, and in fact the responses above seem to indicate that they didn’t.
But I did change the phrasing, so as not to claim I was trying to shut down discussion. The point I was making, however, remains.
I am shocked to hear that people need proof something is an infohazard before deciding that the issue needs to be discussed BEFORE posts like this go live.
I think there’s a few issues here:
When deciding to take down a post due to infohazard concerns, what should that discussion look like?
How thorough should the vetting process for posts be before it gets posted, especially given infohazard and unilateralist curse considerations?
Is this post an infohazard and if so how dangerous is it?
My previous comment was with regards to 1.
With regards to 2, it’s a matter of thresholds. Especially on this forum, I’d want people to err heavily on the side of sharing information even if it might be dangerous. I wouldn’t want people to stop themselves from sharing information unless the potential infohazard was extremely dangerous or a threat to the continued existence of this community. This is mainly due to the issues that crop up once you start blinding yourself. As I understand it, 2 people discussed this issue before posting, and deemed it worthwhile to post anyway. To me, that seems like more than enough caution for the level of risk that I’ve seen associated with this post. Granted, I don’t think the authors took the unilateralists curse into account, and that’s something that everyone should keep in mind when thinking about posting infohazards. It would also be nice to have some sort of “expanding circles of vetting” where a potentially spicy post can start off only being seen by select individuals, then people above a certain karma threshold, then behind a login by the rest of the LW community, and then as a post viewable and linkable by the general public.
With regards to 3. AFAIK, the issue became politicized the second the president issued a press conference claiming the CDC was doing a good job. This prompted every political actor set against the current presidency to go out of their way to find everything wrong with the CDC and the governments overall response to the pandemic. They then started shouting their findings from the rooftops.But this is only after much larger forums than LW have been questioning the response of both the CDC and the WHO for quite a while now.
The cat left the bag a month ago, and in a significantly less levelheaded and more memetically virulent form. This post won’t change that, and is at worst a concise summary of what has already been said.
1) Yes, this discussion is important, but it should have taken place before a post like this was posted.
2) The standard for something that is admittedly an infohazard can’t be ” the authors themselves think the thing they are doing is a good idea.” And knowing most of the people in this space, I strongly suspect that if anyone in biosecurity in EA had read this, they would have said that it needs at least further consideration and a rewrite. Perhaps I wrong, but I think it is reasonable to ask someone in the relevant area, rather than simply having authors discuss it.
People have many avenues of vetting things before making a public post—it’s not like the set of people in EA who work on biosecurity is a secret, and they have publicly offered to vet potential infohazards in other places.
3) I disagree with you that most of the criticism in other places was of the form exhibited here. Claiming we shouldn’t trust the CDC seems dangerous to me. And this post isn’t simply repeating what others have said. Liberal news sources often note that the Trump administration isn’t trustworthy, or say that the CDC has screwed up, but as far as I have seen, they *don’t* claim that the organization is fundamentally untrustworthy, as this post does. In my view, the central framing changes the tone of the rest of the points from “the CDC doesn’t get everything right, and we should be cautious about blindly accepting their claims” to “the CDC is fundamentally so broken that it should be actively ignored.” While I’m sure that some lesswrong readers are going to marginally and responsibly update their beliefs in light of the new information presented here based on their well-reasoned understanding of the US government and its limitations, many readers will not.
Would your opinion change significantly if we changed the wording to highlight that this is an opinion on the trustworthiness of the CDC in this moment, with these constraints, rather than a fundamental property of the CDC?
I read that exactly the opposite way—it says they discussed it, not that they consulted anyone external, much less that they checked with people and were told that there was a consensus that this would be fine. Unilateralist curse isn’t solved by saying “I thought about it, and talked with my co-author, and this seems OK.”
I downvoted because I felt that this:
Was outside of LW norms. It came off as a blunt attempt to shut down discussion, with very little in terms of justification for doing so. This is in no way a clear cut infohazard, and even if it was, I’m not convinced that shutting down discussion of things that might be infohazards is a good policy, especially on a relatively obscure site centered around truth seeking. Statements this confident about issues this complicated should only be said after some extensive analysis and discussion of the situation. jtm’s presentation of the issue struck me as far more tempered and far less adversarial. I’d encourage Davidmanheim to supplant his comment with a more fleshed out version of his position.
I am shocked to hear that people need proof something is an infohazard before deciding that the issue needs to be discussed BEFORE posts like this go live. I see no evidence that any such discussion occurred, and in fact the responses above seem to indicate that they didn’t.
But I did change the phrasing, so as not to claim I was trying to shut down discussion. The point I was making, however, remains.
I think there’s a few issues here:
When deciding to take down a post due to infohazard concerns, what should that discussion look like?
How thorough should the vetting process for posts be before it gets posted, especially given infohazard and unilateralist curse considerations?
Is this post an infohazard and if so how dangerous is it?
My previous comment was with regards to 1.
With regards to 2, it’s a matter of thresholds. Especially on this forum, I’d want people to err heavily on the side of sharing information even if it might be dangerous. I wouldn’t want people to stop themselves from sharing information unless the potential infohazard was extremely dangerous or a threat to the continued existence of this community. This is mainly due to the issues that crop up once you start blinding yourself. As I understand it, 2 people discussed this issue before posting, and deemed it worthwhile to post anyway. To me, that seems like more than enough caution for the level of risk that I’ve seen associated with this post. Granted, I don’t think the authors took the unilateralists curse into account, and that’s something that everyone should keep in mind when thinking about posting infohazards. It would also be nice to have some sort of “expanding circles of vetting” where a potentially spicy post can start off only being seen by select individuals, then people above a certain karma threshold, then behind a login by the rest of the LW community, and then as a post viewable and linkable by the general public.
With regards to 3. AFAIK, the issue became politicized the second the president issued a press conference claiming the CDC was doing a good job. This prompted every political actor set against the current presidency to go out of their way to find everything wrong with the CDC and the governments overall response to the pandemic. They then started shouting their findings from the rooftops. But this is only after much larger forums than LW have been questioning the response of both the CDC and the WHO for quite a while now.
The cat left the bag a month ago, and in a significantly less levelheaded and more memetically virulent form. This post won’t change that, and is at worst a concise summary of what has already been said.
I think these are all good points.
1) Yes, this discussion is important, but it should have taken place before a post like this was posted.
2) The standard for something that is admittedly an infohazard can’t be ” the authors themselves think the thing they are doing is a good idea.” And knowing most of the people in this space, I strongly suspect that if anyone in biosecurity in EA had read this, they would have said that it needs at least further consideration and a rewrite. Perhaps I wrong, but I think it is reasonable to ask someone in the relevant area, rather than simply having authors discuss it.
People have many avenues of vetting things before making a public post—it’s not like the set of people in EA who work on biosecurity is a secret, and they have publicly offered to vet potential infohazards in other places.
3) I disagree with you that most of the criticism in other places was of the form exhibited here. Claiming we shouldn’t trust the CDC seems dangerous to me. And this post isn’t simply repeating what others have said. Liberal news sources often note that the Trump administration isn’t trustworthy, or say that the CDC has screwed up, but as far as I have seen, they *don’t* claim that the organization is fundamentally untrustworthy, as this post does. In my view, the central framing changes the tone of the rest of the points from “the CDC doesn’t get everything right, and we should be cautious about blindly accepting their claims” to “the CDC is fundamentally so broken that it should be actively ignored.” While I’m sure that some lesswrong readers are going to marginally and responsibly update their beliefs in light of the new information presented here based on their well-reasoned understanding of the US government and its limitations, many readers will not.
Would your opinion change significantly if we changed the wording to highlight that this is an opinion on the trustworthiness of the CDC in this moment, with these constraints, rather than a fundamental property of the CDC?
I might be misunderstanding you, but doesn’t Elizabeth explicitly state that this discussion did take place here?
I read that exactly the opposite way—it says they discussed it, not that they consulted anyone external, much less that they checked with people and were told that there was a consensus that this would be fine. Unilateralist curse isn’t solved by saying “I thought about it, and talked with my co-author, and this seems OK.”
Right, so you set the standard higher than simply talking about it. That wasn’t clear to me from your previous post but it makes sense.