Talk:LessWrong

Archives (Index)

2024/April 2024/July 2024/August 2024/September 2025/January 2025/March 2025/April 2025/July 2025/October

This page is archived by ClueBot III.

Quote

@Veilure, I'm a little opposed to including this quote:

The New York Observer noted that "[LessWrong] is fixated on a branch of futurism that would seem more at home in a 3D multiplex than a graduate seminar ... Branding themselves as 'rationalists,' as the Less Wrong crew has done, makes it a lot harder to dismiss them as a 'doomsday cult'."

The quote is old (2012) and topics that were "doomsday cult"-like at the time have now gone pretty mainstream, so this part of the quote didn't age well. And it's not really clear how it fits a "Purpose" section, it's more of a subjective critique of the movement. It seems to me that the existing sentence "LessWrong is also concerned with transhumanism, existential threats, and the singularity." already indicates the weirdness of the movement in a more neutral way.

KingSupernova also objected to that quote in an archived discussion. Alenoach (talk) 18:25, 7 June 2026 (UTC)

The quote is definitely old, but so is the debate on including it in this article. I don't think it's been archived but consensus way back then was to keep it and I don't see how the argument has changed, even with the mainstream-ification of LW's talking points.

If you have more sources that "topics that were 'doomsday cult'-like at the time have now gone pretty mainstream" and can justify why the quote no longer applies, I see no issue with moving it to the past tense, but removing it outright would likely be WP:NPOV in the other direction. Veilure (talk) 07:43, 8 June 2026 (UTC)

It's pretty clear that the topic of existential risks is much less fringe than it was in 2012, see for example this source or that one. But beside that, consensus is needed for inclusion rather than removal (WP:ONUS), and KingSupernova and Secarctangent proposed removing it. Are there more people who expressed support for having it? Old discussions are in the April 2024 archive, I don't think there is any archived discussion missing. Alenoach (talk) 19:08, 8 June 2026 (UTC)

The quote is a bit heavy-handed, but I support including this perspective in some way. The Washington Post source you link cites a poll from Gallup for the bit about AI regulation becoming more mainstream. The poll isn't about extinction risks, it's more focused on AI’s capability to make fair and unbiased decisions. So even if more people are talking about existential risk, saying it's fully mainstream would be too much of a stretch.

The LessWrong view of existential risk aligns more with the AI industry itself, as the NYT source suggests. This version is criti-hype which inflates the importance of AI by implying it's more capable and more powerful than it currently is. Regardless of the intentions of true believers, it also feeds the industry's desire for regulatory capture by derailing proposed legislation.

As for the age of the source, there are a lot of subsequent reliable sources which mention the connection between LessWrong and the Zizians. So if nothing else, this link to "doomsday cults" in 2019 shows that this wasn't a completely unforeseeable consequence of promoting fringe viewpoints like this. Grayfell (talk) 21:40, 8 June 2026 (UTC)

I don't agree with the framing that a lot of LessWrong criti-hypes by exaggerating AI performance, and I think the interest of companies is rather in deregulation, but I don't think it really matters here.

The WaPo article cites the poll for the statement "Surveys already indicate that most Americans support government rules for AI." It's not exclusively about "AI’s capability to make fair and unbiased decisions", it broadly covers AI safety regulation, but even if it were only about fairness, that is just one sentence in the WaPo articles, and it wouldn't refute the broader fact that concern about existential risk from AI is much less fringe than it was in 2012.

The connection with the Zizians isn't really about LessWrong's "Purpose" and is covered in LessWrong#User_base. Alenoach (talk) 05:58, 9 June 2026 (UTC)

Discourse about "AI risk" conflates fringe discussions of existential risk from AGI taking over and killing everybody, with current, non-hypothetical harms caused by AI. OpenAI and Anthropic have been talking about 'existential risk' from their founding, but they would rather people stop talking about deepfakes, worsening wealth inequality, CSAM, suicidal ideation, data centers, etc. Being seen as getting ahead of "AI risk" without having to actually do anything is to the benefit of these companies, regardless of how sincere LessWrong's users are in discussing it.

The Zizians are barely covered in that section, and as I said, there are many additional sources which could be added to expand on this connection. My point is that the article shouldn't mislead readers into taking the website's self description as 'rationalist' at face value when reliable sources do not take this at face value. We're not here to promote this forum, we're here to summarize sources. The Observer article is one example of a source which is skeptical of LessWrong's self-description, but the Zizians incident prompted many more sources to look critically at this website. Grayfell (talk) 18:47, 9 June 2026 (UTC)

I agree with Grayfell. The way the Less Wrong nexus of groups uses the term rationalism is highly at odds with Rationalism as a philosophical discipline. These mismatches of statement to definition also dominate how the Less Wrong nexus treats the threat posed by AI. The contemporary discourse surrounding AI have a lot to do with environmental impact from energy consumption, systemic bias in training sets, misinformation and reputational harm from deepfakes and the use of AI tools to break labour movements. These do not align with fears of Roko's Basilisk or of Paperclip maximizer fantasies which are the principal AI safety concerns of the Less Wrong nexus. We need to make sure that this article is clear on how this influential group of bloggers and tech entrepreneurs define things differently from the general public. Simonm223 (talk) 18:58, 9 June 2026 (UTC)