This is a sensationalist video put out by an influential YouTuber who generally creates good science videos but in this case does not do the work of the FHI justice nor its substantial and pioneering achievements. To sweep the rug from under the feet of such crucial researchers at an important inflexion point in the development of global AI regulation is hugely detrimental to those efforts. Perhaps others would like to give some pushback in the YouTube comments as I myself (and some others) have also done:

"Generally, I enjoy your videos Sabine and think you do a great job explaining science to the layperson. This one is beneath you, however, full of sensationalist cheap jibes, and lacking serious analysis of the very important work that went on at the Institute into existential risk and the dangers of AI. The closure of this institute could not have come at a worse time, seeing as we are on the brink of getting a global body on AI regulation together at the UN. Of course the Institute and its leaders made mistakes, and they are suffering from the reputational fallout of the FTX financial fraud caused by someone who called themself an effective altruist. (The philosophy itself is also not without its flaws and is still evolving.) But you really are throwing the baby out with the bathwater with this video. Very disappointing."




Sorted by Click to highlight new comments since:

Could you edit to include a few sentences of summary and/or why you think readers might be interested in viewing this?

Thanks. Done. Didn't have time to earlier.

Have also made my opinion on it clearer in the title, which was previously neutral and minimally descriptive.

Thanks. I'd suggest waiting to post until you have time to write a little bit. Your post can get pushed off the frontpage pretty quickly if it attracts near-zero or negative karma, and posting a link where the context isn't either obvious or provided will generally draw a mixed-to-negative reaction. I think that's fair -- watching a video is a meaningful time commitment, and it's reasonable to expect posters to provide some information to help users decide whether it is an appropriate one to make. Here, I'd also mention that the video author has over 1.25MM subscribers to her channel, to establish that this isn't some ~random person almost no one is listening to anyway.

I thought I was providing a modestly valuable service to EAs by linkposting these things - they seemed relevant to current EA debates, and it's generally good to see what those outside the movement are saying, to prevent it from becoming too inward-looking. But it doesn't really seem to be valued so I'm not going to bother anymore. All these comments are on the level of 'is a linkpost a good thing or not'. No-one is saying what they think about the actual content of the video, which is what would have been interesting to me. I feel alienated. And bored.

What is interesting, though, is that it appears to be an example of 'kill the messenger' syndrome. People don't appear to like negative things being brought to their attention, and instead of dealing with/responding to the problem/content itself (e.g. what this prominent and well-respected YouTuber says about longtermism, the EA movement, Bostrom's early racism and later apology, the conflict between scientists and bureaucrats at FHI, etc.; or the timing of shutting such an institute at a crucial point in the development of AI regulation) they avoid dealing with their discomfort at the problem by deflecting it towards criticising the message-bearer instead.

It's interesting, albeit somewhat dispiriting, to see this psychological principle, known since ancient times, still alive and kicking amongst some 21st members of the EA forum.

I think there are some members who near-flexively downvote criticism . . . and they tend to vote on the earlier side. In contrast, your potential upvoters are probably not going to upvote without either watching the video or at least reading a good summary.

I do think video is often a bigger ask, as people can't really skim it like they can an article. If people don't want to watch, that is their perogative, maybe their loss. As for the meta-commentary, someone asked why you were getting downvotes, which invited that commentary in.

[comment deleted]2

I've noticed that Deborah often linkposts prominent people being critical of EA, and these posts are often downvoted. Why is this? 

On this specific post, I agree with Jason that an explanation of the relevancy would be useful (although personally, the title alone was enough), but I don't think this imperfection justifies the downvotes. 

Personally, I find them useful because it's valuable to see what external critics are saying about EA. This helps me either a) learn from their critiques or b) rebut their critiques. Even if they are bad critiques and I don't think it's worth my time rebutting them, I think I should be aware of them because it's valuable to understand how others perceive the movement I am connected to. I think this is the same for other Forum users. 

This being the case, according to the Forum's guidance on voting, I think I should upvote them. But, judging by the downvotes, others see it differently. If you are one of these other people could you explain yourself? Thanks!   

I believe Deborah edited the title, and we can't blame voters for voting on the title as it stood at the time of their vote.

I didn't vote, but I think it's reasonable to expect someone posting a link to say something of moderate substance about why they think it might be of value to the reader. I don't think "this is bad criticism of FHI" is enough. Mentioning that this person has a wide audience of 1.25MM would be just enough, while summarizing the critique would be better.

Yeah, the old title was enough for me because I'd heard of Sabine, but I think your advice to have a title that provides more context is good, e.g., "Prominent YouTuber with millions of subscribers posts extensive critique of FHI". I agree it's reasonable to expect someone who is posting a link to say something about its relevance. 

However, I don't think it's reasonable to downvote without first checking the relevancy if that checking can be done in seconds (as was the case here). 

There should be a moderate bar for linkposting, as it takes up one of the frontpage slots. People may be downvoting because they see a link post with no body text as a low-effort post, and thus less likely to reflect consideration of the bar.

I think the bar for linkposts is supposed to the same as it is for original posts, no? At least, that’s how I’ve interpreted guidance published by mods on the subject.

If people are downvoting because they’re assuming a linkpost without a summary is low value then that’s a pity. Summaries are encouraged but they aren’t mandatory.

Or maybe down-voters are following Forum guidance perfectly - they’re downvoting because they don’t think it’s valuable for other Forum users to see prominent people publishing critiques. I disagree with this view, so it would be nice to see a defence of it.

Perhaps it would be valuable to have a ‘quick links’ section where this sort of thing could be shared without taking room from the front page? Or the guidance should advise people like Deborah to post this sort of thing in the quick takes section?

Ultimately, voting is an exercise in judgment by voters applying their own standards. I will say that I've seen very short text posts (as opposed to quick takes) get the same treatment.

Where the linkpost is to a video, I think it's usually low value unless there's enough information to enable the reader to make their own decision about whether to use their time to view it. I'm a little more forgiving with linkposted text, which can be quickly skimmed.

(Again, I did not vote and can only speculate on why others did)

"Personally, I find them useful because it's valuable to see what external critics are saying about EA. This helps me either a) learn from their critiques or b) rebut their critiques. Even if they are bad critiques and I don't think it's worth my time rebutting them, I think I should be aware of them because it's valuable to understand how others perceive the movement I am connected to."

This explains my reasons for linkposting exactly. Perhaps the forum designers could improve the rating concept of linkposts accordingly. It needs to be different from other posts. There needs to be a way for readers to be able to indicate, say:

(1) Thanks for posting, glad it was brought to my attention OR This is irrelevant to EA concerns, why are you posting here? (2) Like/Don't like - the actual content

I was/am not aware of any 'quick takes' section of the forum.

[comment deleted]1
Curated and popular this week
Relevant opportunities