The Test Stand

Is effective altruism evil? Trey Goff published an essay arguing that it is. His friend Niklas Anzinger, founder and CEO of Infinita, thinks that's too strong, and in the first episode of The Test Stand he takes the essay apart.

Niklas runs Infinita, which works to get biotech treatments to early clinical evidence faster, starting in Montana under SB 535, a law that builds on right-to-try. He hosts the Stranded Technologies Podcast, has been a Próspera resident for four and a half years, and his company formed Montana's first ETRB, a private review board.

Trey goes first. He gives effective altruism its strongest case, then makes his own: the malaria bednets that ended up as fishing nets, and how refusing to discount the future puts an infinity in the ethical math and makes AI risk outweigh everything else.

Niklas, a self-described weak deontologist (someone who holds that rights bind you unless the stakes are overwhelming), says effective altruism is possibly not good but not evil, a word he reserves for ideologies like Nazism and communism. He would have written the essay about AI doomerism instead, and he puts the real problem in the impartiality premise: the demand that you weigh strangers as heavily as your own family and children.

Trey gives ground. On the essay's section about moral concealment, he tells Niklas "this is probably the weakest point that I made." Niklas gives some back: he agrees there is a streak of cowardice in effective altruism, and that entrepreneurship, not RCTs and aid programs, is what actually lifts people up.

Then it gets bigger than EA. Why randomized controlled trials are weaker evidence than they look (exclusion criteria, professional trial volunteers, third variables nobody modeled). Why SpaceX's habit of testing hardware until it breaks beats NASA's plan-everything approach. And how someone with radical goals should behave in Washington, including why Trey has found honesty to be the best strategy there, against everything he was told to expect.

If either of them got something wrong, say so. That is the point of the show.

Chapters

00:00 Why this show exists

01:44 Who Niklas Anzinger is

05:05 Trey's steelman of effective altruism

09:34 The bednets that became fishing nets

11:26 Utilitarianism, longtermism, and infinity

17:35 Niklas on weak deontology and "evil"

22:31 AI doomerism as the real danger

25:47 Animal suffering and impartiality

30:04 The drowning child

31:27 Moral concealment and open-source AI

40:59 Trey concedes his weakest point

46:39 Scientism and the limits of RCTs

54:22 The hidden third variable

1:01:32 Cowardice and entrepreneurship

1:08:02 SpaceX versus NASA

1:14:00 Abolitionism versus incrementalism

1:21:28 Decision strategy and tit for tat

1:28:54 Honesty as the winning strategy

1:36:01 Play the game you're good at

1:43:18 Close

Links

Trey's essay, "Effective Altruism is evil": https://www.treygoff.com/writing/effective-altruism-is-evil

Infinita: https://infinita.co

Montana ETRB: https://montanaetrb.org

Niklas on X: https://x.com/NiklasAnzinger

Bednets used as fishing nets (Short et al., PLOS One, 2018): https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0191519