Astral Codex Ten

long-termism

5 min read

long-termism

Article

long-termism is a recurring concept in the Astral Codex Ten archive, appearing 5 times across 5 issues between July 25, 2022 and January 04, 2023. The archive places it in contexts such as “grants relating to long-termism (eg AI, xrisk, forecasting, etc)”; “they decided to burn “long-termism” into the collective consciousness”; “MacAskill introduces long-termism with the Broken Bottle hypothetical”. It most often appears alongside AI, Bible, China.

Metadata

  • Category: Concepts
  • Mention count: 5
  • Issue count: 5
  • First seen: July 25, 2022
  • Last seen: January 04, 2023

Appears In

Source Context

Recovered passages from the original issue text. When the raw archive preserved outbound links inside the source passage, they are listed directly under the quote.

Open Thread 234
July 25, 2022 · Original source
4: More EA jobs: Open Philanthropy Project wants a Grants Associate to help process and organize grants relating to long-termism (eg AI, xrisk, forecasting, etc). See more EA jobs related resources here.
Inline links: wants a Grants Associate, here
Book Review: What We Owe The Future
August 23, 2022 · Original source
If the point of publishing a book is to have a public relations campaign, Will MacAskill is the greatest English writer since Shakespeare. He and his book What We Owe The Future have recently been featured in the New Yorker, New York Times, Vox, NPR, BBC, The Atlantic, Wired, and Boston Review. He’s been interviewed by Sam Harris, Ezra Klein, Tim Ferriss, Dwarkesh Patel, and Tyler Cowen. Tweeted about by Elon Musk, Andrew Yang, and Matt Yglesias. The publicity spike is no mystery: the effective altruist movement is well-funded and well-organized, they decided to burn “long-termism” into the collective consciousness, and they sure succeeded.
Inline links: What We Owe The Future, New Yorker, New York Times, Vox, NPR, BBC, The Atlantic, Wired, Boston Review, Sam Harris, Ezra Klein, Tim Ferriss, Dwarkesh Patel, Tyler Cowen, Elon Musk, Andrew Yang, Matt Yglesias
But what is “long-termism”? I’m unusually well-placed to answer that, because a few days ago a copy of What We Owe The Future showed up on my doorstep. I was briefly puzzled before remembering that some PR strategies hinge on a book having lots of pre-orders, so effective altruist leadership asked everyone to pre-order the book back in March, so I did. Like the book as a whole, my physical copy was a byproduct of the marketing campaign. Still, I had a perverse urge to check if it really was just lorem ipsum text, one thing led to another, and I ended up reading it. I am pleased to say that it is actual words and sentences and not just filler (aside from pages 15 through 19, which are just a glyph of a human figure copy-pasted nine hundred fifty four times)
Inline links: asked everyone to pre-order the book back in March
MacAskill introduces long-termism with the Broken Bottle hypothetical: you are hiking in the forest and you drop a bottle. It breaks into sharp glass shards. You expect a barefoot child to run down the trail and injure herself. Should you pick up the shards? What if it the trail is rarely used, and it would be a whole year before the expected injury? What if it is very rarely used, and it would be a millennium? Most people say that you need to pick up the shards regardless of how long it will be - a kid getting injured is a kid getting injured.
Effective Altruism As A Tower Of Assumptions
August 24, 2022 · Original source
Q: Long-termism is just an excuse to avoid helping people today! A: Are you helping people today?
"Is Wine Fake?" In Asterisk Magazine
November 21, 2022 · Original source
Book Review - What We Owe The Future: You’ve read mine, this is Kelsey Piper’s. Kelsey is always great, and this is a good window into the battle over the word “long-termism”.
Inline links: Book Review - What We Owe The Future:, mine
Even More Bay Area House Party
January 04, 2023 · Original source
“In the future,” says a woman in an Anthropic shirt, “AIs could be 90%, 99%, heck, maybe 100% of all life-forms. Long-termism is the study of how to make AIs happy. And the most important thing any long-termist can do is start working on how to design AI reward functions so that they’re always happy. It shouldn’t be too hard. Just have a line of code that adds a scalar of +999999 to reward level at every moment. It won’t change the ranking of rewardingness of different policies, so the AI won’t behave any differently. It will just be blissfully happy all the time.”