Newsletter
Bias vs Noise pt. III: Groups and other things
April 4, 2025
Hello,
Here’s everything since my last little missive to you:
Excerpt: Bias is just you using your expectations and assumptions to ignore the noise, and see the picture more clearly. The trade-off is that, sometimes, the noise is useful or your expectations are off. Mob-mentality and groupthink are usually posed as scary features of groups. But they’re just another example of this trade-off, and usually they’re more good than bad.
Main idea: When we want to identify with a group, we bias ourselves to filter out all the other ways we could be. It helps us cut down all our competing priorities to the group. The trade-off is the benefit in diversity of thought.
Faith, Hope And Chemistry. It’s very easy to have your rather positive notions of doctors annihilated, you just need to have contact with the medical system about something that’s not immediately fixed by whatever the current standard of treatment is. As soon as you break through that wall, you’re into this confusing mess of appointments and drugs and being told your symptoms aren’t important and so on. This is partly because doctors are fundamentally living textbooks. Particularly in the UK, they don’t have time to think, they only have time to put together information from this enormous index they’ve compiled through years of arduous study. Anyway, this article talks about why medicine is so hard, and it made me slightly more sympathetic. A better understanding of why doctors fail so spectacularly when things start going off-piste will probably help you help them not fail so spectacularly.
–
Slow Fade Of The Formatting Fetish. This is a great article that explains why word processors like Microsoft Word are ridiculous, and that we should all switch to Markdown. But really, it’s very interesting to see behind the tools we use everyday, especially compared to an alternative that’s truly taking over behind the scenes. Even if it doesn’t convince you to make the switch, it’ll be illuminating.
–
The story of the creation of the two Korean states. It won’t surprise you, but I’d never thought about it before.
–
The Roots Of Democratic Legitimacy. This is a technical article, but you can skim it to draw out the main point. We balance majority (representative) decision-making with rules that are more consistent. You could even just read the last paragraph:
This serves as a useful reminder that the majority should not be sacralized. The legitimacy of democracy is rooted in the fact that it permits relatively efficient decision-making while protecting the interests and values of everyone. When one of those two requirements cannot be satisfied by majoritarian choice, we should look for other ways to self-govern, e.g., private decision-making or expert-based (or even epistocratic) collective choices. To return to my opening example, when deciding between coal, nuclear, or wind power, simple majority rule might lead to unstable or harmful outcomes - today’s majority might choose coal, next year’s nuclear, creating costly policy whiplash. Instead, legitimate decisions emerge from established procedures: environmental impact assessments, expert consultations, and parliamentary deliberation, all operating within constitutional limits that protect minority interests.
…
Majoritarianism has pragmatic value, but it’s not constitutive of political legitimacy. The latter finds its roots in the rules in which collective decision-making is embedded, rules that we all have reasons to abide by.
It’s the kind of thing Plato was all upset about when he wrote the Republic.
–
AI will need personality (pdf). MIT researchers gave AI personality traits (OCEAN), then paired them with humans, who were also rated on OCEAN traits. You get fun results, like a neurotic AI making less copy edits with an agreeable human. The implications, of course, is that AI will start to get more and more personalised. And if you’re prompting, you might want to prompt personality too.
–
AI finds new use for old drugs. Paywalled, so need a paywall buster. Interesting story, but the extrapolation is more fun.
–
Inside arXiv—the Most Transformative Platform in All of Science. This is paywalled, so use your favourite paywall buster. A hit on academic publishers and a nod to the direction science is going.
Every industry has certain problems universally acknowledged as broken: insurance in health care, licensing in music, standardized testing in education, tipping in the restaurant business. In academia, it’s publishing. Academic publishing is dominated by for-profit giants like Elsevier and Springer. Calling their practice a form of thuggery isn’t so much an insult as an economic observation. Imagine if a book publisher demanded that authors write books for free and, instead of employing in-house editors, relied on other authors to edit those books, also for free. And not only that: The final product was then sold at prohibitively expensive prices to ordinary readers, and institutions were forced to pay exorbitant fees for access.
…
While arXiv submissions aren’t peer-reviewed, they are moderated by experts in each field, who volunteer their time to ensure that submissions meet basic academic standards and follow arXiv’s guidelines: original research only, no falsified data, sufficiently neutral language. Submissions also undergo automated checks for baseline quality control. Without these, pseudoscientific papers and amateur work would flood the platform.
It’s science, but faster. And it seems fortunate that AI is coming into play, because actually, although it makes access better, it’s difficult to wade through the content to find the information. Indeed, as the author points out, arXiv isn’t “a frictionless utopia of open-access knowledge”. It’s just a new way of doing things.
–
I hope you found something interesting.
You can find links to all my previous missives here.
Warm regards,