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: The best EA Global yet? (And other updates.), published by Lizka on February 13, 2022 on The Effective Altruism Forum.

Events enable attendees to make new connections, learn about core concepts, share and discuss new research, and coordinate on projects.

Summary

We think that EA Global London was our best event yet — probably more valuable than all events we ran in 2019 (combined).

We’re expanding our team so that we can run three large EA Globals, and support at least six EAGx conferences.

You can help us by:

Coming to our events and meeting and advising people who are newer to the community

Pointing strong potential hires our way (email [email protected])

Applying for support to run community events, or encouraging others to apply

Metric: connections

Our core metric is the number of connections made at our conferences. We calculate connections by multiplying the number of attendees times the average number of self-reported new connections. We define a new connection as “someone you feel comfortable reaching out to to ask for a favor,” so this is a relatively high bar.

This metric has a variety of limitations. We’re still fine-tuning it, finding better ways of measuring it, and thinking through other important outcomes to measure.

YearTypeEstimated total new connectionsNotes2019Final8,938Excluding EAGx2020Final15,521Including EAGx2021Q4 only10,700EA Global: LondonFinal17,108Excluding the Coordination Forum, EAGx

EA Global: London

We ran the EA Global: London conference from October 29th to October 31st.

Outcomes

We had 1530 attendees in total (~900 attended in-person, and 630 attended virtually). We hosted 22 different sessions at the in-person event, a number of which were streamed or paired with a session for virtual attendees.

This event created approximately 10,700 connections for the community, which was more successful than our goals.

On average, in-person attendees formed 10.24 new connections. Virtual attendees only formed 2.41.

The feedback has been overwhelmingly positive, especially for the in-person side of the event.

On average, in-person attendees rated their likelihood to recommend (LTR) the event as a 9.1/10 (8.1/10 for virtual attendees)

The net promoter score was 73 for in-person, 29 for virtual attendees.

We doubled the size of the in-person event at very short notice, and were worried that this would cause more logistical chaos or COVID-related issues. Our impression is that this did not happen, and the feedback we got supports this impression.

We think this decision added around 3,900 connections.

The virtual side of the event underperformed, although it still added around 1,500 connections; it was positively received by attendees, but we did not optimize it.

We are reconsidering running hybrid conferences in the future, and might run separate virtual and in-person conferences in order to give more attention to each side.

We began providing grants for travel and accommodation for attendees who otherwise couldn’t attend, particularly students. We also continued to offer discounted tickets for attendees who need them.

Bonus image:

Postponing EAGxPrague

We also helped the EAGxPrague team organize their event in December and then respond to the decision to postpone it as a result of the Omicron wave. We expect that a lot of the work that went into organizing EAGxPrague is not totally lost, since the event will still be happening on 13 - 15 May this year.

Hiring

In Q4, we hired

Eli Nathan, Project Manager for EA Global, starting in April

Ivan Burduk, Events Associate, started in November

Charlotte Potts, a part-time contractor working on revamping our volunteer process

We are currently hiring for:

Events Generalists

This role could involve a wide range of tasks including public communications, impact analysis, admissions processing, ...

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: Knowing, published by LoganStrohl on February 13, 2022 on LessWrong.

I’m going to draw some practical distinctions among types of knowledge, as an attempt to tap into your intuitions and avoid having to give some convoluted, ivory-tower definition of the word. I request that you try not to get distracted by where I’ve drawn my distinctions—the precise placement of the borders is not the point of the exercise. The point of the exercise is to shine your attention on the richness, depth, and complexity of your capacity to know—that the word “know” means more than one thing.

Let’s begin with a little formative assessment.

Do you know what comes out of your kitchen faucet?

Do you know what glaciers are made of?

Do you know what those big white fluffy things in the sky are?

If so, then you are familiar with water. When someone talks about it, you’re not completely lost.

Do you know at what temperature water boils?

Do you know the atomic composition of an ordinary water molecule?

Do you know what percentage of your body’s volume is water?

If so, then you know some facts about water. Your concept of water contains (at minimum) a few isolated pieces of accurate information.

Do you know the way water sounds when it pours into a cup?

Do you know how water feels when it runs over your skin?

Do you know the look of a stream's surface as it glitters in the sun?

If so, then you are able to identify water when you encounter it in real life. You have direct, experiential data. You are able to predict how various encounters with water will impact your senses, and you probably recognize water when those sensations occur (at least sometimes).

Do you know what happens if you leave a beer bottle in the freezer overnight?

Do you know how a water mill grinds grain into flour?

Do you know why it rains?

If so, then you probably have at least one model of water.

Whether or not that model is explicit, it includes enough structure that you can predict the behavior of water in various situations, even if those situations are outside of your own direct experience. Your model might be rudimentary, in which case the above questions probably produced hesitation or “sort of?” and you’d maybe only be able to produce a short paragraph in response to each. Or your model might be rich and deep, in which case your “yes” was confident, and you could in principle write multiple essays on the subject.

Do you know how to swim?

Do you know what to expect when applying watercolor paints to a wet canvas?

Do you know how to make sea water safe to drink?

If so, then you have some practical mastery of water. It’s not just that you recognize water, or that you know some things about it, or that you can predict its behavior—your models of water are integrated with your models of yourself and other parts of the world, accurately and deeply enough that when you personally interact with water in real life, things tend to go more or less as you intend (at least in certain kinds of situations).

Breaking the format of the pop quiz now, to ask a more difficult question:

Can you name other things you know as intimately and thoroughly as water? Is there some swath of the territory with which you have extensive familiarity, lots of factual knowledge, rich predictive and explanatory models, and also practical mastery in a wide variety of situations? In other words, where do you think you might have deep mastery?

One such domain that’s likely common: many people have this portfolio of knowledge when it comes to driving cars. The average American spends a little under an hour a day in a car, so if you’re like the average American in this respect (and also you’re the one doing most of the driving for your household), then you’ve plausibly spent over three thousand hours behind the wheel in the past ten years. And if ...

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: Intro to Naturalism: Orientation, published by LoganStrohl on February 13, 2022 on LessWrong.

A note on how to approach this sequence:

If you were exactly like me, I would ask you to savor this sequence, not scarf it. I would ask you to approach each of these essays in an expansive, lingering, thoughtful sort of mood. I would ask you to read them a little bit at a time, perhaps from a comfortable chair with a warm drink beside you, and to take breaks to make dinner, sing in the car, talk to your friends, and sleep.

These essays are reflections on the central principles I have gradually excavated from my past ten years of intellectual labor. I am a very slow thinker myself; if you move too quickly, I expect we’ll miss each other completely.

There’s a certain kind of thing that happens when a person moves quickly, and relies a lot on their built-up structures—their familiar, tried-and-true habits of thought and perception. There is a different kind of thing that happens when a person can step back and bring those very structures into view, rather than standing atop them. I'm hoping for the latter.

But since you’re not exactly like me, there might be a better way to approach this sequence, in your particular case, than the exact one I’d suggest to myself. I hope you’ll take a moment to check.

What matters to me is not how fast you read, or how many sittings it takes; what matters is that you create for yourself enough space to explore, to observe the real world beyond all these words, to watch how your own thoughts and experiences unfold in dialog with mine. Any method that allows you to maintain that kind of space as you read is perfect, as far as I’m concerned.

“Naturalism” is a label for a conceptual framework, investigatory discipline, and semi-formalized way of looking at and learning about the world. I’ve been developing and teaching naturalism for the past couple of years, if you start counting on the day I chose the term, or since 2013, if you take a more historical perspective. I’ve made some relevant content available, but I’ve had trouble writing a straightforward introductory post.

The reason for this, as far as I can tell, is that the naturalist perspective is suspicious of categories, projections, and preconceptions, and seeks to move closer toward (relatively) unfiltered, direct observations. It’s specifically a frame-breaking and frame-escaping discipline, so it’s hard to describe in frame-terms without being importantly misleading.

I ardently desire not to mislead anyone.

There’s a saying I like a lot, which goes: “A man with one watch knows what time it is; a man with two is never sure.”

(When I first heard this, I needed to pause for a moment, to let it sink in. It helped me to actually visualize wearing a watch on each wrist, then checking the time.)

The reason I like this saying is that it reminds me to be confused, in an appropriate fashion. “Confused” might even be too weak of a word—it’s almost like it reminds me to be scared, in an appropriate fashion.

I mean, sure—for most things, I don’t have to know what time it actually is, with sufficient precision that the off-ness of my watch makes a meaningful difference. The claim here is not that absolute clarity is required at all times.

But there is indeed an unfortunate property of having-a-watch, which is that it provides me with an answer to the question “what time is it?”

It provides that answer clearly, and specifically, and unambiguously. It provides that answer with more confidence than it ought to, like a calculation that doesn’t attend to significant digits. And if I’m not careful, then with my watch right in front of me, it’s very easy to lose track of the fact that I do not, in fact, know exactly what time it is. To forget that what I really know is what time it almost is.

This is what our...

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: D&D.SCP: Anomalous Acquisitions, published by aphyer on February 12, 2022 on LessWrong.

STORY (somewhat longer than usual, still skippable)

When you graduated top of your class from Data Science School, you didn't care where you ended up working, you just wanted to find the highest-paid job possible. (Your student loans may have had an impact on this decision).

(Yes, I know a past scenario background has also begun this way. It's a versatile backstory. And one that I think may resonate with many of my players.)

Somewhat surprisingly, at the last minute Goldman Sachs was outbid by a firm you'd never heard of before called "Marshall, Carter and Dark Ltd." They hired you (for a substantially higher salary than anyone else was offering) to analyze sales data for "our ultra-exclusive chain of clubs, facilities, and banking services, catering to the finest gentlemen since 1952".

You've been working for them for several years now, in the course of which you've seen a variety of...very strange objects pass through their firm. It seems that, in addition to their upper-class business, they acquire and sell a variety of items that don't strictly seem to obey the laws of physics.

In the course of the last few years, you have excelled beyond your employers' expectation at all the technical aspects of your job...while perhaps performing less well on the social aspects.

It might have been turning down an invitation to drinks with your boss Mr. Alden because you don't like alcohol...

...or spending forty-five minutes explaining a particularly interesting piece of your data methodology to a client who had asked an innocuous question (how were you supposed to know he would be such a philistine as to find that boring?)...

...or asking a few too many questions about the various strange items that passed through the firm...

...or (what seemed based on your colleagues' reactions to be worst of all) pointing out that when you acronym "Marshall, Carter and Dark" to "MC&D" it looks very like "McDonald's".

Whatever the reason, when you were called into the office of Mr. Marshall for a private interview, you assumed you were going to be fired.

(You weren't too worried. You're a data scientist. You can find another job in less time than it'll take you to leave the building. And given the severance package in your contract, firing you would basically amount to paying you a large bonus.)

With your usual level of social acumen, as soon as the door to Mr. Marshall's office closed you asked him: "Am I being fired?"

He looked surprised to hear you say that (which would have fooled you three years ago, but since then you've learned that the top-tier client-facing staff of MC&D change expressions the way you change clothes.) "Fired? Of course not! Mr. Alden speaks very highly of your technical ability."

"My technical ability yes, my non-technical ability...uh...not so much." (You're sure a salesperson would have said that in a better-sounding way).

Marshall puts on a smile that you couldn't imitate if you spent months in front of a mirror. "Quite so! In fact, that's why you're here." He steeples his hands and leans forward slightly. "It's come to our attention that you're being...under-utililized in your current role. As you say, your salesmanship skills are...perhaps a point for future improvement. However, your technical skills are top-notch, and we think they're being wasted in your current position. And as it happens, we have a recent opening as head of Acquisitions that my colleagues and I think would benefit from a more data-driven approach."

You blink in confusion. It sounds like they're...promoting you? That can't be right...can it?

"I'm pleased to inform you that this position would come with a 20% base salary increase over your current compensation, with the additional potential for a noticea...

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: Abstractions as Redundant Information, published by johnswentworth on February 13, 2022 on The AI Alignment Forum.

What is it that makes the concept of “pencil” a good abstraction?

One way to frame it: there are lots of “pencils” in the world - lots of little chunks of the world which I could look at which all contain information about pencils-in-general. Many different places from which I could gain information about “pencils”, and many different places where I can apply that information to make predictions. If I don’t know that pencils are usually made of wood with a graphite core, there are many different chunks of the world which I could look at to gain that information. If I do know that pencils are usually made of wood with a graphite core, there are many different chunks of the world where I could apply that information to predict the materials from which something is made.

Alternatively, consider a gear, spinning in a gearbox. What makes the gear's rotational speed a good abstraction? Well, there are lots of little patches of metal comprising the gear. If I don't know the gear's rotation speed, I could precisely estimate it from any of the little patches. If I do know the gear's rotation speed, I can use it to predict the rotation of many different little patches of metal within the gear.

This seems like an intuitively-reasonable notion of what makes abstractions “good” in general: there are many different places from which we can learn the information, and many different places where we can apply the information to make predictions. In other words, a good abstraction is highly redundant: it appears in many different places, and in any of those places we can use the abstract information to make predictions and/or gain more information about the abstraction in general.

In this post I’ll sketch out one way to formalize this idea mathematically, and show that it’s equivalent to the formalization of abstraction as information-at-a-distance. In particular, in the gear example: conditional on the highly redundant information (i.e. the overall rotation of the gear), the low-level rattling of far-apart chunks of metal is statistically independent. More generally, redundancy yields the same high-level abstract information as the Telephone Theorem, but in a mathematically-cleaner form, and without the need for different summaries between different variables.

Meta

Advice for readers: I try to keep the more-dense math in a few specific sections and the appendices, which can all be skimmed/skipped if you just want a conceptual picture.

Epistemic status: I’m aiming for physics-level rigor here. The proofs involve some shenanigans with infinite limits, and I expect them to contain subtle errors. However, I also expect that the results will accurately predict reality in practice, and that whatever subtle errors the proofs contain can be patched by the sort of mathematician who enjoys dealing with tricky limit shenanigans.

Thankyou to Rob Miles and Adam Shimi for suggesting I try Excalidraw.

Basic Idea: Redundant Information Is Conserved Under Resampling

Suppose I sample the genomes of two random humans, G1 and G2.What information is redundant across these two random variables?

One intuitively-reasonable answer: if I threw away one of the two sequences, then the redundant information is whatever I could still figure out from the other. More generally, if I have a whole bunch of genomes from random humans, G1.GN, the redundant information is whatever I could still figure out after throwing one of them away.

To formalize “what I could still figure out after throwing one away”, we’ll use the idea of resampling: I throw away the value of Gi, and then sample a new value for Gi, using my original probability distribution conditioned on all the other genomes. So, for instance, I thr...

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: Risks from Asteroids, published by finm on February 11, 2022 on The Effective Altruism Forum.

When thinking about risks from space, you’ll likely think of comets and asteroids.

The asteroid that caused a mass extinction event approximately 66 million years ago collided with an energy roughly ten billion times as great as the bomb dropped on Hiroshima. Far more significant than the damage caused directly by the impact, it caused a cloud of ash and dust to block out the Sun’s light across the globe — eventually rendering three quarters of the world’s species extinct.

As recently as 1908, an asteroid exploded over a region of remote Siberia, flattening an estimated 80 million trees over more than 2,000 km2 of forest. There is no impact crater; instead the meteoroid likely disintegrated kilometers above the ground in an ‘air burst’ — powerful enough to throw people to the ground as far as 40 miles from the centre of the explosion. If a similar-sized asteroid instead exploded over a large metropolitan area, it might have killed more people in a single day than any event in history.

Worse still, an asteroid greater than 10km in diameter colliding with Earth could massively disrupt Earth’s climate, likely causing a long ‘winter’ making it far harder to grow crops. The result could be the premature deaths of most living people, human progress thrown back by decades, or perhaps — given some mechanism by which survivors would fail to repopulate — permanent extinction.

So wouldn’t the space programs of the world do well to urgently unite around building effective asteroid defence systems? The answer isn’t so clear.

First, we know that catastrophic asteroid impacts must be very infrequent, and asteroids that pose significant existential risks like extinction are even rarer.

Without even trying to count asteroids in the night sky, we can look at the Earth for indications of large craters or extinction events, and reason that asteroids over 1km in diameter can’t strike much more frequently than once every million years on average — any higher and the relatively small number of craters wouldn’t make sense.

Of course, we can also try to spot asteroids in the sky. Astronomers have identified a large majority of near-Earth asteroids larger than 1km across, and many smaller examples. From these surveys, we know that the chance of an Earth-impact for asteroids 1-10km in diameter is about 1 in 6,000, and about 1 in 1.5 million for asteroids larger than 10km across — that is, roughly the size of the asteroid that caused the Cretaceous–Paleogene mass (dinosaur) extinction event.

But we can be even more confident about the risk from asteroids in this century specifically, because astronomers can track the large asteroids they spot and check whether their trajectories are on course to intercept with Earth’s. With this more precise information, the risks look lower still. In particular, the chance of an impact from an asteroid greater than 10km in size following its natural orbit is effectively zero if NASA is correct in claiming it has identified all asteroids this size.

In The Precipice, Toby Ord summarises the risks:

Asteroid sizeTotalFoundAverage CenturyNext Century1-10 km~ 920~ 95%1 in 6,0001 in 120,00010 km~ 4> 99%1 in 1.5 million< 1 in 150 million

Spotting and tracking asteroids is important: knowing whether they pose a risk this century means knowing more about how best to allocate our resources across different kinds of risk. Very likely, we needn’t worry about building asteroid defence until the deflection technology is cheaper, more effective, and easier to govern; but there’s a slim chance we may need an all-out effort on deflection now because we spot an asteroid heading our way. Clearly, it pays to know which world we’re in.

Fortunately, the international community has already d...

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: Simplify EA Pitches to "Holy Shit, X-Risk", published by Neel Nanda on February 11, 2022 on LessWrong.

TL;DR If you believe the key claims of "there is a >=1% chance of AI causing x-risk and >=0.1% chance of bio causing x-risk in my lifetime" this is enough to justify the core action relevant points of EA. This clearly matters under most reasonable moral views and the common discussion of longtermism, future generations and other details of moral philosophy in intro materials is an unnecessary distraction.

Thanks to Jemima Jones for accountability and encouragement. Partially inspired by Holden Karnofsky’s excellent Most Important Century series.

Disclaimer: I recently started working for Anthropic, but this post entirely represents my opinions and not those of my employer

Introduction

I work full-time on AI Safety, with the main goal of reducing x-risk from AI. I think my work is really important, and expect this to represent the vast majority of my lifetime impact. I am also highly skeptical of total utilitarianism, vaguely sympathetic to person-affecting views, prioritise currently alive people somewhat above near future people and significantly above distant future people, and do not really identify as a longtermist. Despite these major disagreements with some common moral views in EA, which are often invoked to justify key longtermist conclusions, I think there are basically no important implications for my actions.

Many people in EA really enjoy philosophical discussions and debates. This makes a lot of sense! What else would you expect from a movement founded by moral philosophy academics? I’ve enjoyed some of these discussions myself. But I often see important and controversial beliefs in moral philosophy thrown around in introductory EA material (introductory pitches and intro fellowships especially), like strong longtermism, the astronomical waste argument, valuing future people equally to currently existing people, etc. And I think this is unnecessary and should be done less often, and makes these introductions significantly less effective.

I think two sufficient claims for most key EA conclusions are “AI has a >=1% chance of causing human extinction within my lifetime” and “biorisk has a >=0.1% chance of causing human extinction within my lifetime”. I believe both of these claims, and think that you need to justify at least one of them for most EA pitches to go through, and to try convincing someone to spend their career working on AI or bio. These are really weird claims. The world is clearly not a place where most smart people believe these! If you are new to EA ideas and hear an idea like this, with implications that could transform your life path, it is right and correct to be skeptical. And when you’re making a complex and weird argument, it is really important to distill your case down to the minimum possible series of claims - each additional point is a new point of inferential distance, and a new point where you could lose people.

My ideal version of an EA intro fellowship, or an EA pitch (a >=10 minute conversation with an interested and engaged partner) is to introduce these claims and a minimum viable case for them, some surrounding key insights of EA and the mindset of doing good, and then digging into them and the points where the other person doesn’t agree or feels confused/skeptical. I’d be excited to see someone make a fellowship like this!

My Version of the Minimum Viable Case

The following is a rough outline of how I’d make the minimum viable case to someone smart and engaged but new to EA - this is intended to give inspiration and intuitions, and is something I’d give to open a conversation/Q&A, but is not intended to be an airtight case on its own!

Motivation

Here are some of my favourite examples of major ways the world was improved:

...

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: We need more nuance regarding funding gaps, published by Joey on February 12, 2022 on The Effective Altruism Forum.

TLDR: Funding gaps should be talked about in more specific ways, e.g., considering specific cause areas, organizational scale, and funder diversity. When broken down this way, there are likely far more gaps than many EAs imagine. Our model should look more like the table below than a simple “whether EA has strong funding or not” aggregate.

Funding gaps are a well-discussed topic in the EA movement and between funders of all sizes. The concept overall is highly relevant and useful when making decisions. However, I think the bulk of communication about funding gaps is quite unspecific and unrefined, and even carefully communicated content can lead to confusion in overall discourse. A claim like “EA has a funding gap” or “EA does not have a funding gap” is too unspecific a heuristic and can lead to a lot of confusion about the state of funding, as some strong charities receive no funding despite a consensus that the given area is considered “funding flooded.” Working with a number of early-stage charities going through the Charity Entrepreneurship Incubation Program has really given me a more nuanced sense of where specific gaps are and how important this information can be.

Cause area variation

Funding gaps quite clearly change across cause area, and it could be true overall that AI has very few funding gaps while mental health has many. But I think the level of nuance could even go a lot deeper than that. Corporate campaigns in animal welfare might be well-funded while, at the same time, funding in vegan outreach is quite limited. Funding differences according to location can also be radical, with a project in London getting far more funding than the same project based out of Abuja. Most funders and EAs recognize this, although it's sometimes forgotten in the broader discourse. A less-considered factor is that organizations might have other differentiating factors outside of cause area that affect their funding options. This possibility is what I want to delve into more in-depth for this post.

Organizational size variation

A factor that is not often talked about is that the size (and correlating factors like age) of an organization can be just as influential to funding availability as cause or intervention area. For example, there are areas with large governmental funders who only consider organizations of a certain size and age or have requirements that would be near impossible for a smaller organization to fulfill. Governments are not the only entities with these restrictions – large funders are often looking for large opportunities. Some funders are uniquely keen on very large megaprojects but, of course, relatively few of them launch from scratch, with most experiencing a slow build-up of resources and capacity. Example: global povertyTo take global poverty as an example, GiveWell is typically looking for organizations that can absorb considerable amounts of funding in the near future (e.g., 10 million per year or more). This strategy is well thought out.

GiveWell wants to move a huge amount of funding and only has time to evaluate a certain number of organizations to the level of depth they require in order to recommend them. It would take 10x as much evaluation time to consider 10 organizations that can each only absorb 1 million per year. Although this strategy makes sense in GiveWell’s case, it can have a strange effect on the ecosystem. It may result in small projects not getting funding because of the idea that GiveWell would already have funded strong projects in the area (without a donor necessarily knowing whether a charity was ruled out for cost-effectiveness or simply size). It also means there are considerable impacts that may be left on ...

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: If you're after a job in the US: the H-1B Lottery is in six weeks, published by Andy Jones on February 12, 2022 on The Effective Altruism Forum.

There is a lot of impact to be had working in the US, what with US organisations having easier access to capital than most anywhere else in the world.

If you're looking to join a US-based organisation, you'll need a US work visa. The common employer-sponsored visas are, roughly,

O-1 'extraordinary talent' visa, which requires you to check three of these eight boxes,

Specific-national visas, like for Canadians, Australians, Chileans and Singaporeans,

H-1B visas, which have laxer requirements but are subject to a 1-in-3-ish lottery for for-profit orgs. It's uncapped for non-profits!

If you're not one of the listed nationals and you don't check the O-1 boxes, the H-1B is the next best option. The lottery happens only once a year, at the end of March, and if you win it's another six months before you can start work.

This means that if you want to work for a US organisation and you're happy to try your luck, it's a good idea to start interviewing at those organisations around about now.

(This is written on the basis of only my own experience of the US immigration system; you'll want to do a lot more research while evaluating your own options. The exactness of the H-1B timing came as a surprise to me, and maybe this'll help other folk)

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: Doubts about Track Record Arguments for Utilitarianism, published by Erich Grunewald on February 12, 2022 on The Effective Altruism Forum.

Utilitarians occasionally submit as evidence in favour of their philosophy the fact that early utilitarians were advocates for social and political positions that have now become norms. One example among many, from the authors of utilitarianism.net:

While not constituting an argument per se, it is worth noting that utilitarian moral reasoning has a strong track record of contributing to humanity's collective moral progress. [...] As a progressive social reformer, Jeremy Bentham defended issues such as the separation of church and state; the abolition of slavery and of capital punishment; legal regulations to protect criminals and non-human animals from cruel treatment; and the decriminalization of homosexuality. [...] John Stuart Mill defended the provision of social welfare for the poor and of freedom of speech. He was the second MP in the UK Parliament to call for women's suffrage and advocated for gender equality more generally. [...] Henry Sidgwick advocated for women's education and the freedom of education from religious doctrines.

They conclude that, though early utilitarians "were still far from getting everything right, their utilitarian reasoning led them to escape many of the moral prejudices of their time and reach more enlightened moral and political positions". (They do write that this is not an "argument per se". I'm not sure what they mean by that. The OED says that an argument is "a reason or set of reasons given in support of an idea, action or theory". The track record as presented does seem like a reason given in support of utilitarianism, so I will call it an argument in this post, with the reservation that I may be missing some subtlety.)

Or here in Peter Singer's words, from this interview:

I think utilitarianism [...] is actually a reforming impulse. [...] Bentham and [later] utilitarians have been against slavery, they've been for women's rights. They've been for the rights of gay people long before anybody else dared to even talk about that. They've been against cruelty to animals. They've been for prison reform. There's a long list of things that utilitarians have been trying to reduce the amount of suffering in relation to and I’m very happy to be part of that tradition and to think of utilitarianism not merely as something for philosophers to talk about, but something that motivates people to act.

(And he is too modest to mention it, but of course we can add to that illustrious list Singer's own work on animal welfare and global poverty.)

This is a somewhat different version, where the track record doesn't show that utilitarianism is correct, exactly, but that it's useful – that it produces good outcomes. But it has in common with the first that the track record is given as a reason in support of utilitarianism.

Summary

I think it's true that early utilitarians were better on social and political issues than e.g. Kant and his followers. However,

I think this (implied) difference is exaggerated (70% confidence); and

I think this tells us very little about the soundness of the utilitarian philosophy (75% confidence).

I base (2) mainly on three points:

The various interpretations of the track record argument that I could think of don't seem very convincing.

There are potentially confounding variables, in particular the culture that early utilitarians grew up in.

The sample size is very small.

Why the Difference between Utilitarians and Kantians May Be Exaggerated

Jeremy Bentham (1748-1832), John Stuart Mill (1806-1873) and Henry Sidgwick (1838-1900) are sometimes contrasted with a dour, inflexible Immanuel Kant (1724-1804). (I'm using Kantian ethics here because that is what I know better, but you can pro...