Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Editing Advice for EA Forum Users, published by Justis on April 22, 2022 on The Effective Altruism Forum.

Ok, I fibbed. This is a crosspost of Editing Advice for LessWrong Users, but I suspect there is a decent amount of overlap between the sorts of editing advice most useful for these two forums.

Who am I?

I'm an editor who does feedback requests for LessWrong. Many of these feedback requests are for proofreading/clarity of language, so I have lots of experience reading community draft posts and helping improve them. I want to share some high level takeaways - low effort self-editing techniques that, based on my[1] experience, I think many to most LW posters could benefit from.

Disclaimers

I wrote all the example text in this post myself. I make no claims that any of it is reasonable on the object level - it's just meant to resemble lots of the stuff I edit.

Bolded text in examples is for emphasis of the points under discussion, not a sign that it would make sense to actually bold those words or phrases.

As usual, reverse advice as necessary.

What's the advice?

Here's my main advice, in decreasing order of confidence - I'll elaborate on each:

Beware "this"

Don't let hedging be a tic

Break up run-on sentences

Use more links, images, examples, and commas

Beware "this"

The problem: many sentences on LessWrong begin with "this". At worst, the thing "this" is pointing to is fully ambiguous between multiple live options, making it impossible to parse the author's intended meaning. But even in milder cases, it can waste a few seconds for the reader to trace out what a post is talking about.

The solution: scan through your draft for words like "this" or "that" (but for whatever reason overwhelmingly "this" specifically), and when in doubt about clarity, just replace them with whatever their intended antecedents are.

An example:

Many key figures in the community have been shortening their timelines recently, since major advances just keep coming one after the other. This is especially worrying, especially compared to discussions of "AI winter" just a few years ago.

What, precisely, does "this" point to in the sentence above? The fact of key figures shortening their timelines? The major advances? When we get later in the sentence and see that we're comparing whatever "this" is to "discussions", we can infer that we're talking about the shortening of timelines (vs. the actual technical advances), but it takes up precious reader bandwidth to figure this out, or, worse, the reader becomes overloaded by ambiguity and unconsciously switches to "skimming mode".

Individually, overuse of "this" isn't a big deal. But it adds up. It's the single greatest contribution to a hand-wavy feeling in the posts I review, like maybe the author isn't totally sure what they're saying. And it's easy to avoid! Just ctrl-f "this" and go to town!

Don't let hedging be a tic

The problem: LessWrongers care a lot about making factually accurate claims. This (gasp!) is a virtue. But like any virtue, it's vulnerable to Goodharting: when people learn that hedging is a sign of epistemic responsibility, you end up with a whole lot of hedging. Occasionally, you actually end up with hedging that is logically incoherent or actively redundant.

The solution: Make sure you hedge precisely as much as necessary, and especially make sure you don't hedge the same claim multiple times in the same way. If "you suspect" something "might" happen, it doesn't also have to be "possible" that it "could" have "potentially" significant effects.

An example:

The other salient quality of insect suffering is its sheer scale. Insects are the most populous animals on the planet, so it may be that if their suffering is morally relevant at all, I believe it could swamp almost any other considerations about animal welfare.

Some hedging is...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: How to set up a UK organisation (Limited Company version), published by velutvulpes on April 22, 2022 on The Effective Altruism Forum.

It's less scary than you think to set up a UK organisation.

If you want to set up a UK business that has social, charitable or community-based objectives, you can set up as any of:

limited company

charity, or from 2013, a charitable incorporated organisation (CIO)

co-operative

community interest company (CIC)

sole trader or business partnership

This guide explains how I set up a limited company, to try and make it easier for others to follow the same steps I did.

(1) Form your company (1 hour)

You can use/ to form your company. For a small fee they will set up your limited company for you.

I recommend using the Privacy option. In this option you'll have a business address registered in London, so that it's that address rather than your home address which is registered publicly. Post that is sent to that London business address is scanned and emailed to you.

(2) Get bank accounts and set up accounting (2 hours)

Banking

I recommend using/ as your business bank account

I would also set up a/ account for any international payments you have

I recommend using as company cards to make all purchases on. It has good automatic systems to scan receipts for any purchases that you make, which is useful for accounting later.

Accountants and book keepers

I recommend immediately setting up with these accountants/ once you've registered your company and have banking set up. It's good to be on top of accounting from the beginning.

(3) Learn about your responsibilities (4-8 hours)

This can be done later, over the first few months of running your organisation.

What is a limited company?

Obligations for limited companies

Paying yourself through your business

Bookkeeping for small businesses

Bookkeeping for Limited Companies

Claiming for business expenses

Corporation Tax for Limited Companies

(4) Done!

That's all you need to get started. It's easier than you think.

Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: I Caught Covid And All I Got Was This Lousy Ambiguous Data, published by Elizabeth on April 21, 2022 on LessWrong.

Tl;dr I tried to run an n of 1 study on niacin and covid, and it failed to confirm or disprove anything at all.

You may remember that back in October I published a very long post investigating a niacin-based treatment protocol for long covid. My overall conclusion was “seems promising but not a slam dunk; I expect more rigorous investigation to show nothing but we should definitely check”.

Well recently I got covid and had run out of more productive things I was capable of doing, so decided to test the niacin theory. I learned nothing but it was a lot of effort and I deserve a blog post out of it null results are still results so I’m sharing anyway.

Background On Niacin

Niacin is a B-vitamin used in a ton of metabolic processes. If you’re really curious, I describe it in excruciating detail in the original post.

All B vitamins are water-soluble, and it is generaly believed that unless you take unbelievably stupid doses you will pee out any excess intake without noticing. It’s much harder to build up stores of water-soluble vitamins than fat vitamins, so you need a more regular supply. Niacin is a little weird among the water-solubles in that it gives very obvious signs of overdose: called flush, the symptoms consist of itchy skin and feeling overheated. Large doses can lead to uncontrolled shaking, but why would you ever take that much, when it’s so easy to avoid?

People regularly report response patterns that sure look like their body has a store of niacin that can be depleted and refilled over time. A dose someone has been taking for weeks or months will suddenly start giving them flush, and if they don’t lower it the flush symptoms will get worse and worse.

Some forms of niacin don’t produce flush. Open question if those offer the same benefits with no side effects, offer fewer benefits, or are completely useless.

Niacin And Long Covid

There’s an elaborate hypothesis about how covid depletes niacin (and downstream products), and this is a contributor to long covid. My full analysis is here. As of last year I hadn’t had covid (this is antibody test confirmed, I definitely didn’t have an asymptomatic case) but I did have lingering symptoms from my vaccine and not a lot else to try, so I gave the protocol a shot.

My experience was pretty consistent with the niacin-storage theory. I spent a long time at quite a high dose of the form of niacin the protocol recommends, nictonic acid. My peak dose without flush was at least 250mg (1563% RDA) and maybe even 375mg (2345% RDA). When I hit my limit I lowered my dose until I started getting flush at the new dose, and eventually went off nicotnic acid entirely (although I restarted a B-vitamin that included 313% RDA of a different form). That ended in September or early October 2021. It made no difference in my lingering vaccine symptoms.

In early 2022 I tried nicotinic acid again. Even ¼ tablet (62.5mg, 390% RDA) gave me flush.

I Get Covid

Once I developed symptoms and had done all the more obviously useful things like getting Paxlovid, I decided it would be fun to test myself with niacin (and the rest of the supplement stack discussed in my post) and see if covid had any effect. So during my two weeks of illness and week of recovery I occasionally took nicotinic acid and recorded my results. Here’s the overall timeline:

Day -2: am exposed to covid.

Day 0: test positive on a cue test (a home test that uses genetic amplification).

Lung capacity test: 470 (over 400 is considered health).

Start Fluvamoxine and the vitamin cocktail, although I’m inconsistent with both the new and existing vitamins during the worst of the illness. Vitamin cocktail includes 313% RDA of no-flush niacin, but not nicotinic acid.

Day 1: sy...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Infra-Topology, published by Diffractor on April 22, 2022 on The AI Alignment Forum.

Attention conservation notice: This post has Less Basic Inframeasure Theory as a prerequisite and goes pretty heavy on the topology.

The basic results are that a whole lot of the fiddly details involved with having inframeasures over non-compact spaces automatically pop out of just trying to make everything continuous, and that some of the old results in LBIT are either wrong or about the wrong topology. Accordingly, this post completely obsoletes the section about topology of inframeasures in Less Basic Inframeasure Theory

For the rest of this post, assume X is a Polish space. Basically, a Polish space is a space that has a metric, all Cauchy sequences have limiting points (like R and unlike Q), and there's a countable subset of it s.t. every point can be written as a limit of points from the countable subset (Q is the corresponding countable subset of R).

Proofs available on request because they're 115 pages and this post was 15 pages.

Topologies and Convergence Notions

The first observation to make is that the type signature of an inframeasure over the space X (or, one of the possible type signatures) is CB(X)→R, where CB(X) is the space of bounded continuous functions X→R. You feed a function in, you get an expectation out, simple as that.

So, an extremely natural requirement to impose is that an inframeasure should be a continuous function. The topology on R is obvious. The topology on CB(X), the space of continuous bounded functions X→R, is much less obvious.

Another observation is that different notions of function convergence correspond to different topologies on function spaces.

You could say that fn converges to f if supx|f(x)−fn(x)| limits to 0. This is uniform convergence, The late fn are close to f everywhere, the rate of convergence doesn't depend on anything.

Or you could say that fn converges to f if, for all x, |f(x)−fn(x)| limits to 0. That's pointwise convergence, for every point, the late fn are close to f at that point. The rate of convergence depends on the point you picked.

Or you could say that fn converges to f if, for all compact sets C⊆X, we have maxx∈C|f(x)−fn(x)| limiting to 0. This is compact convergence. For every compact set, the late fn are close to f on that compact set. The rate of convergence depends on the compact set you picked.

All these different notions of function convergence are linked to different topologies on CB(X). The sup-norm topology, the weak topology, and the compact-open topology, respectively, for uniform convergence, pointwise convergence, and compact convergence.

So, what notion of convergence seems to "naturally" show up in a lot of proofs? Does it seem like we're using uniform convergence, pointwise convergence, or something like that? No. The notion of convergence that has showed up basically every single time while grinding through the inframeasure theory has been the following notion of convergence.

Definition 1: Strict Convergence A sequence of bounded continuous functions fn strictly converges to f if, for all compact sets C⊆X, limn→∞maxx∈C|fn(x)−f(x)|=0, and supnsupx∈X|fn(x)∞

So, basically, strict convergence is uniform convergence on compact sets, along with the entire function sequence having some uniform upper and lower bounds.

What's the motivation for this? Well, there's a few ways to do it, but here's a nice one. Let's say we want eval:CB(X)×ΔX→R to be a sequentially continuous function. Ie, if the sequence of probability distributions μn converges to μ, and fn converges to f, the expectation of fn w.r.t. μn should converge to the expectation of f w.r.t. μ. limn→∞μn(fn)=μ(f) This requires two questions. First, what does it mean for μn to converge to μ, and what does it mean to say that fn converges to f?

Well, the ...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Infra-Miscellanea, published by Diffractor on April 22, 2022 on The AI Alignment Forum.

This post is a writeup of some miscellaneous results that don't cohere into a single big post, but might possibly be of interest.

Also, in mostly unrelated news, this stuff can be a bit tricky to learn for the first time. So, I was thinking of assembling an infra version of the fixpoint exercise sheet, to help cross the gap from "There's all these complicated posts, and it looks nifty-but-intimidating" to "I understand the basic concepts well enough to casually use them when applicable, and can work through most of the requisite math on my own". This particular realm of math is one where it seems pretty feasible to guide someone to reinvent large chunks of it on their own with sufficiently leading discussion, and recreating something yourself does wonders for being able to actually wield it elsewhere. So, if anyone wants to chip in with assembling the problem sheet, or testing it out by spending a day trying out various math problems, please send me a message.

Result Summaries:

1: Because the vector space of signed measures on X and the vector space of continuous functions X→R are dual to each other, inframeasures are dual to infrafunctions! An infrafunction can be viewed as either a concave functional mapping a measure to an expectation value to expectation values, or as a set of a-functions. Compare this to how an inframeasure can be viewed as either a concave functional mapping a function to an expectation value, or as a set of a-measures. The theory of these sorts of things is highly incomplete at the moment.

2: Convex spaces are spaces with a distinguished function ΔX→X, which takes the average of a mixture of points. R being a convex space (kinda) is why the notion of "average of a distribution" exists. So, we can ask "what sorts of space X have a distinguished function □X→X?", to get the infra-analogue of the notion of "average". As it turns out, it's a perfect halfway point between probability theory, and lattice stuff/set theory. They're convex spaces that are also complete lattices, and the probabilistic structure must interact with the order structure in the appropriate way. Specifically, given a probability distribution over subsets, mixing the sets together and taking the infinimum should produce the same result as mixing the infinima of the sets together.

3: What is the space of inframeasures over the empty set? It's just fun working through this, in the same way as it's fun working out why minimizing over the empty set produces infinity.

4: Besides the semidirect product and the free product, another notion of product has shown up. In the specific case of crisp infradistributions, it's the closest match to the naive notion of the product of probability distributions, as it lets you factorize expectations in the same way as products of probability distributions do.

5: The notion of cross-entropy of crisp infradistributions is defined, rounding out the trio of "entropy, cross-entropy, KL-divergence", where we previously had two of the three. An equality in the probability-theory case (entropy plus KL-divergence is cross-entropy) is shown to be an inequality here. Cross-entropy has a nice intuitive interpretation.

6: There's a function fix:[X→□X]→□X (where X is a compact space) that gets you the Kleene fixpoint theorem (from computer science), the Banach fixpoint theorem, and that one theorem about Markov chains having a stationary distribution.

7: A classic result of probability theory is the disintegration theorem, which roughly states that given a probability distribution ν:Δ(X×Y), it can always be written as being produced by a probability distribution μ:ΔX, and a Markov kernel k:X→ΔY giving you the conditional probabilities. This is the theorem that secretly under...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Are smart people's personal experiences biased against general intelligence?, published by tailcalled on April 21, 2022 on LessWrong.

TL;DR: Collider between g and valued traits in anecdotal experiences.

IQ tests measure the g factor - that is, mental traits and skills that are useful across a wide variety of cognitive tasks. g appears to be important for a number of outcomes, particularly socioeconomic outcomes like education, job performance and income. g is often equated with intelligence.

I believe that smart people's personal experiences are biased against the g factor. That is, I think that people who are high in g will tend to see things in their everyday life that suggest to them that there is a tradeoff between being high g and having other valuable traits.

An example

A while ago, Nassim Taleb published the article IQ is largely a pseudoscientific swindle. In it, he makes a number of bad and misleading arguments against IQ tests, most of which I'm not going to address. But one argument stood out to me: He claims that IQ tests only tap into abilities that are suitable for academic problems, and that they in particular are much less effective when dealing with problems that have long tails of big losses and/or big gains.

Essentially, Taleb insists that g is useless "in the real world", especially for high-risk/high-reward situations. It is unsurprising that he would care a lot about this, because long tails are the main thing Taleb is known for.

In a way, it might seem intuitive that there's something to Taleb's claims about g - there is, after all, no free lunch in intelligence, so it seems like any skill would require some sort of tradeoff, and ill-defined risks seem like a logical tradeoff for performance at well-defined tasks. However, the fundamental problem with this argument is that it's wrong. Nassim Taleb does not provide any evidence, but generally studies on IQ don't find g to worsen performance, and tend to find it to improve performance, including on complex, long tail-heavy tasks like stock trading.

But what I've realized is that there might be a charitable interpretation for Taleb's argument. Specifically, we have good reason to believe that his claim is a reflection of his personal experience and observations in life. Let's take a look:

Who is Nassim Taleb?

Nassim Taleb used stock trading methods that were particularly suited to long tail scenarios to become financially independent. With his bounty of free time, he's written several books about randomness and long tails. Clearly, he cares quite a lot about long-tailed distributions, for good reasons. But perhaps he cares too much - he is characteristically easy to anger, and he spends a lot of time complaining about people who don't accept his points about long tails.

Based on this, I would assume that he quite heavily selects for people who care about and understand long tails when it comes to the people he hangs around. This probably happens both actively, as he e.g. loudly blocks anyone who disagrees with him on Twitter, and probably also passively, because people who agree with him are going to be a lot more interested in associating with him than people who he blows up at.

But in addition to selecting for people who care about long tails, I would assume he also selects for smart people. This is a general finding; people tend to associate with people who are of similar intelligence to themselves. It also seems theoretically right; he's primarily known for "smart topics" like long tail probabilities, and his work tends to be about jobs that are mostly held by people in the upper end of the IQ range.

Simulating collider bias

Selecting for both intelligence and black swan awareness exposes Taleb to collider bias: If you select for multiple variables, you tend to induce a negative correlation b...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Covid 4/21/22: Variants Working Overtime, published by Zvi on April 21, 2022 on LessWrong.

Earlier this week, a judge overturned the transportation mask mandate, which I covered here. I will also be covering the situation in China in its own post.

No one asked this time around, but a friend recommends this P100 mask if you’re looking to go big rather than go home.

The biggest other news is that Omicron continues to cycle into new variants.

There’s BA.2.1.12, there’s BA.4 and there’s BA.5.

All seem to be more infectious than BA.2.

Which was more infectious than BA.1.

Which was much more infectious than Delta.

Which was more infectious than Alpha.

Which was more infectious than the original strain.

We’ve come a long way. China is still attempting Zero Covid, which I will cover in the China post, but that is looking more and more like state failure to adapt to reality than a reasonable policy choice. Covid-19 is here to stay, at least until we can get a nasal vaccine that prevents transmission, research towards which is not currently something we are even willing to properly fund.

Executive Summary

Moderna booster candidate looks great.

Cases and deaths radically down on paper due to Easter, mostly ignore.

Things increasingly returning to normal, travel mask mandates lifting.

China situation is scary, will be covering that in its own posts.

Let’s run the numbers.

The Numbers

Predictions

Prediction from last week: 250,000 cases (+13%) and 3,900 deaths (-10%).

Results: 231,576 cases (+4%) and 2,402 deaths (-45%!).

Prediction for next week: 277,800 cases (+20%) and 4,100 deaths (+71%!?)

Easter (and perhaps Passover) suppressed cases and somehow even after two years I seem to not properly adjust for such things – but even then, I mean, that death number, come on. When I break things up into regions it seems likely that last week wasn’t getting ahead of itself, and we should expect some ‘catch up’ growth in cases this coming week.

The death count is clearly artificially low, at least in terms of comparing it to expected future calculations, especially in the South. I am anticipating many of those missing deaths being counted in the coming week. There is however also the possibility that this also partly represents the drop-off of ‘legacy’ deaths that are only sort-of Covid deaths, as it is still above 1% of cases reported a few weeks ago whereas the real IFR is going to be a lot lower than that. So it’s not like the number is fully impossible.

I will note that there has been no mention of the decline in deaths among all the people talking about the transportation mask mandate being lifted. Now more than ever, this is not about death.

Deaths

A number of states reported zero deaths. A lot of others had epic drops of more than half.

Cases and deaths in the UK have diverged, which many say is due to the decline in testing.

In such situations, the place to look is the positive test percentage.

This confirms that the undercount got larger during March before peaking, and there was likely a similar undercount to what we saw in the January peak. Yet the number of cases recorded was far lower while the number of deaths was briefly higher. I do not think the death count was ever actually higher than it was in January, but despite that this is still suggesting that the test percentage stat was underestimating the degree of undercounting.

Saying that case counts ‘no longer have value’ seems to me like going too far. It definitely has some value, and especially it has value in relative terms now that we have a baseline for the new policy regime in the UK. What we can’t do is directly compare free-test world to not-free-test world without making the right adjustments.

As always, if you wait long enough, deaths tell the story. The story they are telling is that things in the UK are roughly...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Are deference games a thing?, published by Daniel Kokotajlo on April 18, 2022 on LessWrong.

Everyone starts off with a voting card. People can give away their cards. At the end, people cast votes using the cards they have, so people with lots of cards get lots of votes. (If you don't like what your card is being used for, you always have the right to take it back.)

Maybe this could be used to implement something more like direct democracy? Have lots of votes every day instead of just one every four years. Most people don't vote on most issues, instead they give their card to their wisest friend who gives it to their wisest friend who has hundreds of cards and spends some free time each day thinking about what to vote on. Other people (e.g. Oprah, Elon Musk) would have millions of cards and would have staffs of advisors telling them how to vote. They'd basically be like congresspeople.

Surely there's a literature on this idea. What terms should I search for?

Also: Is it a good idea? How would it be worse than the current system?

Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Reflections on My Own Missing Mood, published by Conor Sullivan on April 21, 2022 on LessWrong.

Life on Earth is incredibly precious. Even a tiny p(DOOM) must be taken very seriously. Since we cannot predict the future, and since in a debate we should be epistemically humble, we should give our interlocutors the benefit of the doubt. Therefore, even if on my inside view p(DOOM) is small, I should act like it is a serious concern. If the world does end up going down in flames (or is converted into a giant ball of nanobots), I would not want to be the idiot who said "that's silly, never gonna happen."

The problem is my emotional state can really only reflect my inside view and my intuitions. To account for the very real probability that I'm wrong and just an idiot, I can agree to consider, for the purposes of discussion, a p(DOOM) of say, 0.25, which is equivalent to billions of people dying in expected value. However, I can't make myself feel like the world is doomed if I don't actually think it is. This is the origin, I think, of missing moods.

For me personally, I have always had an unusually detached grim-o-meter. Even if an unstoppable asteroid were heading to Earth and certainly we all were going to die, I probably would not freak out. I'm just pretty far out on that spectrum of personality. Given that, and given that it makes me a bit reckless in my thinking, I'm very grateful that the world is full of people who do freak out (in appropriate and productive ways) when there is some kind of crisis. Civilization needs some people pointing to problems and fervently making plans to address them, and civilization also (at least occasionally) needs people to say "actually it's going to be okay." Ideally, public consensus will meet in the middle and somewhat close to the truth.

I've had three weeks to reflect on EY's post. That post, along with the general hyper-pessimism around alignment, does piss me off somewhat. I've thought about why this pessimistic mood frustrates me, and I've come up with a few reasons.

Crisis Fatigue

Here is a highly scientific chart of current world crises:

My younger sister is quite pessimistic about climate change. Actually scratch that: it's not climate change that she's worried about, it's biosphere collapse. I'm conflicted in my reaction to her worries. This is strikingly similar to my internal conflict over the alignment crisis. On the one hand, I'm very optimistic about new energy technologies saving the day. On the other hand, biosphere collapse is just about the most serious thing one could imagine. Even if p(Dead Earth) = 0.01, that would still be a legitimate case for crisis thinking.

With so many crises going on today, and with a media environment that incentivizes hyping up your pet crisis, everyone is a bit fatigued. Therefore, our standards are high for accommodating a new crisis into our mental space. To be taken seriously, a crisis must be 1) plausible, 2) not inevitable and 3) actually have massive consequences if it comes to pass. In my opinion, AGI and the biosphere threat both check all three boxes, but none of the other contender crises fit the bill. Given the seriousness of those two issues, the public would be greatly benefited if all the minor crises and non-crises that we keep hearing about could be demoted until things cool off a bit, please and thank you.

To be clear, I'm putting AGI in the "worth freaking out about" category. It's not because of crisis fatigue that I'm frustrated with the pessimistic mood around alignment. It's because...

Pessimism is not Productive

Given crisis fatigue, people need to be selective about what topics they engage with. If a contender crisis is not likely to happen, we should ignore it. If a contender crisis is likely to happen, but the consequences are not actually all that serious, ...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: My experience with imposter syndrome — and how to (partly) overcome it, published by Luisa Rodriguez on April 21, 2022 on The Effective Altruism Forum.

This is a linkpost from/

In a nutshell: I have imposter syndrome, and I suspect many other people trying to have a big impact with their career do, too. For a long time, imposter syndrome massively limited the impact of my career by causing me to rule out potentially impactful options and affecting my day-to-day job performance and satisfaction — but through talk therapy and other tools, I’ve mostly overcome it. If you think you might struggle with imposter syndrome, tackling it may be one of the most important forms of personal development you can do. It was for me.

I’ve felt like an imposter since my first year of university.

I was accepted to the university that I believed was well out of my league — my ‘stretch’ school. I’d gotten good grades in high school, but I’d never seen myself as especially smart: I wasn’t selected for gifted programmes in elementary school like some of my friends were, and my standardised test scores were in the bottom half of those attending my university.

I was pretty confident I got into the university because of some fluke in the system (my top hypothesis was that I was admitted as part of an affirmative action initiative) — and that belief stayed with me (and was amplified) during the decade that followed.

Throughout that decade, there was evidence that I really was good at my work at different points, but I could always come up with an explanation for why the evidence was unreliable.

For example, as an undergraduate, I was the only first-year student in my biology department to get a research internship at the Mayo Clinic — one of the most prestigious biomedical institutions in the US. But I felt I only got the internship because I’d met the right person at the right time, and tricked them into thinking I was smarter than I was by saying smart-sounding things.

Likewise, during my final year of university, I was given an award for being the top performer in my sociology department. But I believed that was just because the professors ‘liked me’ and thought I was ‘nice.’

My feelings didn’t end at university. They came back when I was offered an internship at GiveWell after graduating. But despite the fact that I had completed multiple work tests and had interviewed with the CEO, I told myself I only got that internship because my resume made me look more competent than I was (because of years of tricking other people into giving me internships and awards!).

More recently, I published a series of blog posts on nuclear war on the Effective Altruism Forum while working at Rethink Priorities, and got a lot of positive feedback. But I also got some negative feedback about some mistakes I’d made in my analysis — which, to me, confirmed that anyone reading my work closely would realise the work was riddled with mistakes and flawed reasoning.

And then when Will MacAskill, apparently impressed by my nuclear war research, asked me to do research into civilisational collapse for his upcoming book, I felt like even more of a fraud — like I’d tricked Will into thinking I was a good researcher by publishing low-quality research that ‘passed’ as good.

Rather than assuaging my insecurities, my successes actually made me feel more ashamed, because they meant that I was fooling more and more people in increasingly high-stakes roles. Now that I was being asked to do things that I thought really mattered, I was going to really be letting people down.

Writing this now, I’m astounded by the mental gymnastics I was performing to maintain the belief that I was fooling everyone. Looking back, I’m proud of the achievements in that list. And even long before I was working on those impactful projects — and even if...