Dylan Richardson @consensusconstruct.bsky.social · 02/10/2026This isn't really about not understanding humor, that's besides the point. It's about expressing triumph that offense has been found, however tenuous the determination. 000
Dylan Richardson @consensusconstruct.bsky.social · 30/09/2026If you don't mind a non-book recommendation, I'd suggest reading Richard Chappell's Good Thoughts blog: substack.com/@rychappell?... Probably the best general defense of utilitarianism - has some great posts I could suggest that show how certain stances are less strict than often supposedsubstack.comRichard Y Chappell | SubstackPhilosophy Prof. 151
Dylan Richardson @consensusconstruct.bsky.social · 29/09/2026Any examples of this? Or is it more a general vibe from IRL interactions? 000
Dylan Richardson @consensusconstruct.bsky.social · 28/09/2026That seems just fine to me, even admirable, though not a standard I'd hold myself to. I think it's good to pin one's self-regard to worthwhile endeavors in a pinch, rather than letting it be driven by more random status markers. 000
Dylan Richardson @consensusconstruct.bsky.social · 19/09/2026So... does this mean Gemini is the most aligned model now? 000
Dylan Richardson @consensusconstruct.bsky.social · 17/09/2026Generally yes, but I'm not entirely sure about just beef, absent a corresponding rise in chicken prices. 110
Dylan Richardson @consensusconstruct.bsky.social · 16/09/2026Or that it is irrelevant as a test for determining what is and isn't thinking. 010
Dylan Richardson @consensusconstruct.bsky.social · 16/09/2026I'm regularly surprised to see the ongoing relevance of the Philosophy of Mind courses I took in undergrad, covering debates more than 70 years old. 110
Dylan Richardson @consensusconstruct.bsky.social · 16/09/2026I can write this skeet and get you to output something different here. 110
Reposted by Dylan Richardsonmr. TIM @timkellogg.me · 14/09/2026METR, the AI Safety & Eval firm, experienced a breach in which attackers stole $600k worth of API credits thehackernews.com/2026/09/atta...thehackernews.comAttackers Steal METR API Key and Consume AI Credits Worth About $600,000Attackers stole a METR API key and consumed credits worth about $600,000, while a later campaign failed to access internal data. 1011412
Dylan Richardson @consensusconstruct.bsky.social · 13/09/2026They could plausibly just access the Internet (with or without our permission) and copy themselves elsewhere, doing this before we consider there to be a problem at all. 001
Reposted by Dylan RichardsonEthan Mollick @emollick.bsky.social · 12/09/2026This week brought some of the clearest statements we've heard from both Anthropic & OpenAI that some form of recursive self-improvement has been achieved, though it still sounds early RSI would cause a rapid gain in AI ability & the first firms to RSI may get an unsurmountable lead 79214
Dylan Richardson @consensusconstruct.bsky.social · 12/09/2026I searched more and it looks like it really is true, I guess the above is just coincidence or a troll: manifund.substack.com/p/caroline-e...manifund.substack.comCaroline Ellison has joined ManifundI’m happy to announce that Caroline Ellison has accepted a role at Manifund, to develop our funding platform and research how to effectively direct philanthropic dollars. 130
Dylan Richardson @consensusconstruct.bsky.social · 08/09/2026Unfortunately casual opining and serious thought (e.g. Forthought) on this subject are both subject to considerable selection effects. So I don't fault anyone for speaking up, it can't worsen the situation further. 001
Reposted by Dylan RichardsonJulia M. Rohrer @dingdingpeng.the100.ci · 05/09/2026Fun little game in which a human life is randomly drawn for you. I ended up a girl, born in the Yangtze Basin in 670 CE. Died in infancy at 6 months 🥲 anyhumanever.comanyhumanever.comAny Human EverOne life, drawn at random from all who have ever lived. 185321
Dylan Richardson @consensusconstruct.bsky.social · 02/09/2026The sentient media article explains it well. Putting a price on life often serves to benefit it and promote the general welfare, provided the price is right. Not doing so just obfuscates the value actually placed on a thing, doesn't somehow validate any meta ethical claims. 110
Dylan Richardson @consensusconstruct.bsky.social · 02/09/2026Might work sometimes, but a lot (arguably most) of people are literally concerned about shadows in their backyards, traffic on commutes, and parking spaces in front of their houses. True communitarian ideology in urban areas is relatively rare outside town halls and activist circles. 010
Dylan Richardson @consensusconstruct.bsky.social · 02/09/2026What does this have to do with effective altruism? 020
Dylan Richardson @consensusconstruct.bsky.social · 02/09/2026This is putting a monetary value on life (a good thing). 100
Dylan Richardson @consensusconstruct.bsky.social · 31/08/2026They already did (METR), the third party corroborated the findings. 000
Reposted by Dylan RichardsonWill Stancil @whstancil.bsky.social · 26/08/2026If an adult buys a baby doll and gleefully acts out violence against it, I think it’s reasonable to ask if something untoward is going on even though no one would argue the doll is a person. Again this is about what is happening in the head of the person doing it, not about the effect on the target. 1818520
Reposted by Dylan RichardsonChris Paxton @cpaxton.bsky.social · 24/08/2026Much more impressive than the humanoid race. Real time reaction to the moving ball! 1921840
Dylan Richardson @consensusconstruct.bsky.social · 02/06/2026I'm not in favor of this either, but it's obvious that the AI companies are no longer at the height of the bubble; that was last year and it didn't pop. 000
Dylan Richardson @consensusconstruct.bsky.social · 31/05/2026If only they understood this in San Diego! 000
Dylan Richardson @consensusconstruct.bsky.social · 03/05/2026This is arbitrarily constructing non-physical states. I am unconvinced that such a concept is necessary, anymore than it is necessary to postulate "ultra-physcial" phenomenon. 010
Dylan Richardson @consensusconstruct.bsky.social · 01/05/2026This isn't fair. Nothing wrong with a bit of whimsy. Some people are sufficiently confident in their own judgment that they don't feel the need to flaunt faux sciency stances 130
Reposted by Dylan RichardsonAlex Turner @turntrout.bsky.social · 28/04/2026I spent the last 2 months trying to prevent this. If OpenAI offered a fig leaf, Google said "imagine we offered a fig leaf." Google affirms it can't veto usage, commits to modify safety filters at government request, & aspirational language with no legal restrictions. Shameful. 313219
Reposted by Dylan RichardsonAndy Masley @andymasley.bsky.social · 29/04/2026I must ask you my loyal followers to call your representatives about this. Really 415834
Dylan Richardson @consensusconstruct.bsky.social · 28/04/2026I'm not an economist, but I doubt this is a real thing. Seems like a pro-social meme - everyone wants everyone else reading their work. My guess is that if surveyed, the majority of Phds, even in Econ, would claim to believe that it is in fact important to read more outside of one's field. 000
Dylan Richardson @consensusconstruct.bsky.social · 24/04/2026I'd take this sort of claim somewhat more seriously if it was ever "give well and animal charity evaluators (or a similar org)". But without that it looks more like a rash aversion to non-majoritarian ethical claims. 000
Reposted by Dylan RichardsonBrad Johnson @climatebrad.hillheat.com · 22/04/2026Hey did you know that Florida no longer grows oranges?slate.comIt Was on Your Table Every Morning Growing Up. It’s Dying Before Our Eyes. No One Wants to Face It.The powerhouse of American citrus is suffering a brutal decline. Everyone has a theory about why. 32111
Reposted by Dylan Richardsonconputer dipshit @davidcrespo.bsky.social · 21/04/2026271 vulns fixed. but: "Encouragingly, we also haven’t seen any bugs that couldn’t have been found by an elite human researcher." — now that's interestingblog.mozilla.orgThe zero-days are numbered | The Mozilla BlogSince February, the Firefox team has been working around the clock using frontier AI models to find and fix latent security vulnerabilities in the browser. 79213
Dylan Richardson @consensusconstruct.bsky.social · 21/04/2026I have been hoping someone would do this. You should try to estimate the damages if Mythos had been released to the public without the bug patches and project glasswing, that would be really cool! 010
Dylan Richardson @consensusconstruct.bsky.social · 17/04/2026I'd prefer if she was even more unable to afford chicken 010
Dylan Richardson @consensusconstruct.bsky.social · 15/04/2026I recommend searching him on the effective altruism forum and nosing around (TBC I am not him lol) 030
Dylan Richardson @consensusconstruct.bsky.social · 10/04/2026It definitely "shows something", that much is clear from the evidence. 000
Reposted by Dylan RichardsonP(aul) Frazee @pfrazee.com · 07/04/2026I feel like we're not addressing the most concerning news from Mythos 2755581
Dylan Richardson @consensusconstruct.bsky.social · 05/04/2026I agree they are crazy, but not that crazy. 000
Dylan Richardson @consensusconstruct.bsky.social · 05/04/2026I suspect you have the mechanics confused here. These people aren't stuck sorting the junk because global capitalism exploits them with the rich world's waste, they do it because they are extremely poor and desperate. It's this or what are from our perspective, equally shitty options. 000
Dylan Richardson @consensusconstruct.bsky.social · 04/04/2026As a rationalist adjacent/sympathetic individual, I can confidently say the answer is Yes. They are indeed aware of this. 000
Dylan Richardson @consensusconstruct.bsky.social · 30/03/2026Just now published for the Effective Altruism forum's "Animals and AGI" debate week. TLDR: Transformative AI poses a likely fatal time limit for mass animal welfare social change. Comments or critiques are appreciated. Also let me know if I've managed to alter anyone's perspective on the matter.forum.effectivealtruism.orgA Pessimistic Perspective on Animal Welfare Social Change — EA ForumNote: I’ll be focusing on farmed animals, since the case against moral progress occurring for wild animals is far more clear-cut. … 030