An Average Dialogue

TLDR: 📝📝📝📝📝📝📝📝📝📝📝📝📝

A: Knock-knock!

B: Who's there?

A: Cowgo.

B: Cowgo who?

A: The past year has been a distressingly-active one, for this community. Here, of course, I'm referring to the union of the rationality community, the Effective Altruism community, and the AI alignment/DontKillEveryoneism (aka "DKE") community. A quick timeline in case you've been in cryostasis under a rock: April 1, 2022: MIRI, the OG AI DKE organization, announced its new "Death With Dignity" strategy. This gave the AI alignment/safety field a kick in the ass, which seems pretty good. Importantly, it also increased the percentage of the community that takes DKE seriously. Probably. Hopefully. I haven't run a survey. This announcement was so impactful, it overshadowed the concurrent announcements of Good Heart Tokens (!) and infant outreach (!!!). RETRACTED August/September-ish 2022 (but written about in January 2023): The Future of Life Institute, an organization dedicated to lowering extinction risk, allegedly gave $100,000 to an extreme right-wing online publication connected to FLI President Max Tegmark's brother. Tegmark denied this. EDIT: the money seems to have been a grant that was applied for, but rejected by FLI and never paid out. Yay! H/t to DC for correcting me on this. August 2022: TIME Magazine prints a full cover story about Effective Altruism! Yay! November-ish 2022: One of the largest donors to EA causes, the crypto exchange FTX, goes bankrupt in a sudden, spectacular, and arrest-warrant-inducing fashion. Somehow, despite allegedly being on lots of stimulants, the CEO Sam Bankman-Fried was unable to even play League of Legends properly. Thus, his downfall could be described as part ethical failing and part skill issue. Anyway, this has impacted EA funding in the short- and medium-terms, just due to the sheer volume of money and committed money that the FTX Future Fund was putting towards many causes, including longtermism. December 2022: I graduate college! Yay! January 9, 2023: Nick Bostrom, an OG AI DKE researcher, writes an apology for past racist writing. A dramatized reading of his apology can be watched here. March 14, 2023: GPT-4 is released (under an API/paywall) by OpenAI. This has made a lot of AI DKE researchers very angry and been somewhat regarded as a bad move. (Not the paywall I mean, that part's far better than e.g. open-sourcing the model.) March 15, 2023: TIME Magazine releases a story about how some Effective Altruist organization leaders had been warned about the personal qualities of FTX's leadership. The good news is, some of the above events were good! It's good to be upfront about AI extinction risks, it's good for fraud to get rooted out sooner than later, and it's even somewhat good for people to bring up their own bad behavior quicker, modulo how well they actually go about it. The bad news is more self-explanatory. It's sad that the community has been rocked, rolled, and rattled by these events. It sucks that some of us, who were probably not trying to be evil, made decisions that harmed the world as a whole.

effective altruism is small. the rationality community is small.

You think people are going to get subtly misled by the PhilosophyTube video, or that a turn of phrase in TIME is going to move the needle on the course of history.

Here's what most people, to the extent they think of us at all, are more likely thinking: "Oh, effective altruism? Didn't that crypto guy make that up to hide his crypto scam?" That is IT. That is ALL most people will hear, PERIOD. A TIME cover won't change that. I'm not even sure a Presidential speech would be big enough to give people the "right idea". - my notes for an earlier, angrier draft of this.

Really, I considered making today's post more metacrap, just totally devolving into the insanity and screaming into the void. Ironically, it might've even gotten more updoots. But, well... I still have hope in this community. I still have hope in the fundamental radical ideas, the stuff lower on the assumption-stack, the spirit of reaching higher than what's "socially acceptable". Because the stakes are, frankly, too high not to. The GPT-4 threads on Twitter alone should dispense with the notion that "business as usual" will continue on planet Earth. The continued existence of mass suffering should remind us that "business as usual" was never enough in the first place. When a social group is rocked like this, the temptation is for everyone to defect-defect. To submit to emotions and ditch logic. To turn truth into a luxury and secrets into weapons. Instead of that, let's work towards cooperation on things that we already notice are, or could be, big problems. Not just "work harder for 5 minutes", but "find changes we can make that move the needle on each cause". People will listen.

B: Okay but where's the funny part?

A: Wait, this was a dialogue? What do you mean?

B: Where's the part of this post that makes it worth posting now? Surely you didn't just post it today to get eyeballs on it, right?

A: Mu.