The Nonlinear Library

The Nonlinear Library

By The Nonlinear FundEducation
Download on the App Store

All The Nonlinear Library episodes

  • EA - [JOB] Opportunity to found Charity Entrepreneurship NGO (outside of the incubation program): Tobacco taxation advocacy by Yelnats T.J.
    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: [JOB] Opportunity to found Charity Entrepreneurship NGO (outside of the incubation program): Tobacco taxation advocacy, published by Yelnats T.J. on August 1, 2023 on The Effective Altruism Forum.
    Overview of job posting
    Highlights
    Why found a NGO/charity
    Why tobacco taxation
    The role
    Who your future co-founder is
    The selection process
    How to apply
    FAQs
    Recommended readings
    Charity Entrepreneurship is often called the Y combinator of the charity/NGO space and its alumni are compared to the PayPal Mafia (but for the charity sector).
    Note: this is not an application for the upcoming CE incubation program. You can apply for that here.
    Early application deadline: August 6th 23:59 UTC
    Application deadline: August 13th 23:59 UTC
    After the deadline, applications will be taken on a rolling basis.
    The process is going to be fast and most applicants that go through the whole process will do so in about 3.5 weeks. Times will be quicker for applicants that submit earlier at each stage. Turnaround times will typically be a matter of days.
    Highlights
    The opportunity
    This is a CE-researched intervention with a CE incubatee - J.T. Stanley - that will be considered for seed funding (via the CE seed funder network) in mid-September.
    (The CE incubatee went through the winter 2023 cohort. His original match for co-founder from the cohort ended up pursuing another career opportunity with comparable impact that arose near the end of the program.)
    The impact
    Tobacco is 13x Malaria in deaths (WHO, 2023).
    Whereas malaria, HIV, and neonatal deaths are all decreasing year over year, deaths from tobacco are increasing. Tobacco is currently on track to kill a billion people in the 21st century (WHO, 2021).
    Here is a link to a section that talks about the financial and economic impacts.
    Taxation is both the most effective and most neglected form of tobacco control (WHO; NIH, 2016; World Bank, 2017).
    Multiple EA organizations have ran analysis that found tobacco taxation advocacy to be "an extremely cost-effective intervention."
    CE's cost-effectiveness analysis put the expected value between 39 and 51 USD per DALY averted.[1]
    Open Philanthropy's back-of-the-envelope calculation found 30 USD per DALY averted in expectation.[2]
    Giving What We Can's (GWWC) report suggested that cost per life saved could go as low as 800 USD according to one scientific study.[3]
    In other words, a successful tobacco taxation charity would immediately have an impact with cost-effectiveness at the upper echelons (think Against Malaria Foundation for GiveWell and Lead Exposure Elimination Project for CE).
    Why you should apply
    This is your chance to be a founder of an impactful organization.
    Don't count yourself out. If in doubt apply. Roughly half of the successful applicants/founders that are accepted into the CE incubation program did not think they would qualify. Don't preempt yourself from becoming a founder; let the process play out.
    If you've applied to CE in the past, this is the opportunity for you.
    Why found an NGO/charity
    Pros
    You are not just a cog in the machine; you have ownership of a venture/organization that could deliver massive impact. The health of hundreds of thousands of tobacco users depends on you and your co-founder's fateful actions.
    One of the best expected values for impact - This intervention has the potential to avert a ridiculously high amount of deaths and DALYs.
    Upward career mobility and increased impact later in your career. Running an impactful organization will give you critical skills, experience, and career capital that will better position you for impact later in your career.
    Cons
    A much more serious commitment than your typical job.
    Founder salaries of CE charities are very modest compared to other parts of the EA ecosystem and the first-year salary is lean.
    Founding an organization...
    16 min
  • EA - You Can Also Help Animals By Earning (More) in Other Career Paths and Donating by Animal Advocacy Careers
    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: You Can Also Help Animals By Earning (More) in Other Career Paths and Donating, published by Animal Advocacy Careers on August 1, 2023 on The Effective Altruism Forum.
    "Animal Advocacy Careers" do not need to be only careers in animal advocacy organisations
    This is the first piece of AAC's new "Impactful Animal Advocacy Career Paths" series (more posts coming soon).
    Animal advocacy organisations play a vital role in creating a compassionate world for animals, but they face significant challenges in securing adequate funding. Almost all organisations need financial support in order to sustain and expand their programs. Many of these organisations are funding constrained.
    Providing these resources creates an essential, indirect yet major impact for animals through empowering the organisations helping them.
    Why Donations For Animals Are Crucial
    Almost every advocacy organisation would like to:
    hire more staff who would work on important tasks,
    create more advertisements that would raise awareness about animal rights,
    start more programs to grow and strengthen the movement.
    None of these things are free.
    It is simply unrealistic and unfair to expect individuals to selflessly work for animals without any compensation. It is also hard to attract skilled, and competent individuals to work on animal advocacy without decent salaries.
    There are 3 main problems limiting our progress in putting an end to animal suffering:
    Lack of funding in nonprofits
    Animal advocacy organisations listed a lack of funding as the single most important thing that limited their organisations impact.
    For these reasons, organisations require extensive and stable funding.
    Unless animal nonprofits are adequately resourced, they cannot create meaningful impact for animals. While some organisations also earn money by selling merchandise or event tickets, they simply cannot function like a for profit company.
    Nonprofits primarily depend on donations to cover their expenses.
    Uneven distribution of funding
    The problem is not just the limited amount of money in the animal advocacy space, it is also about how the funds are shared and distributed.
    Farmed and wild animal welfare organisations receive significantly fewer donations compared to organisations working on companion animals, animals used in labs, and captive animals. Yet, the number and suffering of farmed and wild animals are much higher than these other groups.
    Farmed and wild animal advocacy is relatively more neglected than other animal advocacy fields, in terms of both public attention and funding.
    While "for every one dog or cat euthanized in a shelter, 3,400 farmed land animals are confined and slaughtered", farmed and wild animal organisations receive about %1 of total donations to animal charities (which also include shelters, organisations that focus on companion animals, etc.).
    Graphic by Animal Charity Evaluators
    Struggling Against Powerful Institutions
    Finally, the institutions that these organisations are struggling against are extremely powerful. The annual donations to farmed and wild animal advocacy organisations is approximately 200 million dollars.
    Animal agriculture industry, on the other hand, is a billions of dollars worth industry where tens of thousands people work each day.
    The animal advocacy movement, farmed and wild animal advocacy in particular is miniscule in terms of financial power compared to the institutions it is positioned against.
    When animal welfare organisations stage campaigns against corporations like McDonalds or Kroeger in order to achieve institutional welfare reforms, they struggle against organisations that have hundreds, even thousands times more financial resources than they do.
    Is Alt Protein the Solution? Exploring Ambitions and Challenges
    The alternative protein sector has much more investment and mar...
    26 min
  • LW - Exercise: Solve "Thinking Physics" by Raemon
    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: Exercise: Solve "Thinking Physics", published by Raemon on August 1, 2023 on LessWrong.
    Note: please write any answers to this prompt in spoiler-tags.
    Recently I set out to deliberate practice at "reasoning about confusing intellectual problems."
    Eliezer's Class Project has a fictional group of rationality students try to find the true theory of quantum gravity in one month. This always seemed like a cool goal and test for rationality training to aspire to. If you're not solving difficult open problems faster than science, your Art of Rationality probably isn't complete.
    Of course, our Art of Rationality isn't complete yet. But, I think there is something promising in this area, as a way to ground out "rationality training" in something concrete. It seems like good practice to take a given physics question you don't understand the theory behind, and try to invent the theory yourself.
    I don't think we're anywhere close to the "definitively impressive" version of rationality practice/training. But, I think a good next step is "Solve Thinking Physics™"
    Thinking Physics is a textbook teaching physics "question-first" - it presents a physics-y situation, and asks you to figure out what happens next. The questions are multiple choice, but often fairly tricky nonetheless.
    I think a good rationalist-training goal is aim for a goal of "be (correctly) 95% confident in the answer", as a rough proxy for "there were no major lingering confusions about the problem except for generic 'maybe I missed something?'". And, failing that, have the subgoal of at least being calibrated about how confused you. Every time you look at an answer, first log your probabilities for each of the multiple-choices in Fatebook.io (or prediction-tracking tool of your choice).
    The problems are set up in a way that you can probably reason about them from some basic background knowledge, without much math background. They're ideal for people who don't have much physics background (since the whole point of the book is to teach you physics), although I know people with got a physics degree 10 years ago who still find it fairly hard.
    I spent two weeks working on Thinking Physics problems, and hosting meetups/workshops where other people could join me. With each question, I focused on learning as much as I could about how-to-think.
    My original hypothesis was that I could get significantly better at it in 6-8 weeks. I only spent two, and the result so far is I think I'm significantly better although didn't yet hit my goal of 95% accuracy. (In my final test-set, I got 1 out of 5 questions wrong, when I was aiming for zero. I do think I have a pretty clear sense of why I got that 1 question wrong, and what I should have done differently)
    After workshopping some ideas for "the Thinking Physics rationality challenge", I now present you with three tiers of challenge.
    Challenge I: Solve three problems (and learn from them)
    Step 1: Do an exercise.
    Spend some time trying to solve three Thinking Physics question. Aim for 95% accuracy, fully deconfusing yourself about each exercise.
    Write down your probabilities for each answer.
    It's important to actually write down the probability for each answer - otherwise, you may get a vague sense of "yeah that's probably right", that doesn't allow me to cleanly say "I got this one wrong." And doing it for all the answers, not just your favorite one, gives you additional bits about whether your models made any sense. (i.e. having clearly stated "I think answer A is most likely and B is second most likely" gives you a harder update if it turns out that A and B were both wrong)
    Step 2: Learn from it
    Then, think about how you could have solved the problem better.
    Your primary goal is to learn as much as possible from each question.
    Babble as many new insights as you can about how to th...
    8 min
  • LW - The "public debate" about AI is confusing for the general public and for policymakers because it is a three-sided debate by Adam David Long
    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 "public debate" about AI is confusing for the general public and for policymakers because it is a three-sided debate, published by Adam David Long on August 1, 2023 on LessWrong.
    Summary of Argument: The public debate among AI experts is confusing because there are, to a first approximation, three sides, not two sides to the debate. I refer to this as a three-sided framework, and I argue that using this three-sided framework will help clarify the debate (more precisely, debates) for the general public and for policy-makers.
    Broadly speaking, under my proposed three-sided framework, the positions fall into three broad clusters:
    AI "pragmatists" or realists are most worried about AI and power. Examples of experts who are (roughly) in this cluster would be Melanie Mitchell, Timnit Gebru, Kate Crawford, Gary Marcus, Klon Kitchen, and Michael Lind. For experts in this group, the biggest concern is how the use of AI by powerful humans will harm the rest of us. In the case of Gebru and Crawford, the "powerful humans" that they are most concerned about are large tech companies. In the case of Kitchen and Lind, the "powerful humans" that they are most concerned about are foreign enemies of the U.S., notably China.
    AI "doomers" or extreme pessimists are most worried about AI causing the end of the world. @Eliezer Yudkowsky is, of course, the most well-known to readers of LessWrong but other well-known examples include Nick Bostrom, Max Tegmark, and Stuart Russell. I believe these arguments are already well-known to readers of LessWrong, so I won't repeat them here.
    AI "boosters" or extreme optimists are most worried that we are going to miss out on AI saving the world. Examples of experts in this cluster would be Marc Andreessen, Yann LeCun, Reid Hoffman, Palmer Luckey, Emad Mostaque. They believe that AI can, to use Andreessen's recent phrase, "save the world," and their biggest worry is that moral panic and overregulation will create huge obstacles to innovation.
    These three positions are such that, on almost every important issue, one of the positions is opposed to a coalition of the other two of the positions
    AI Doomers + AI Realists agree that AI poses serious risks and that the AI Boosters are harming society by downplaying these risks
    AI Realists + AI Boosters agree that existential risk should not be a big worry right now, and that AI Doomers are harming society by focusing the discussion on existential risk
    AI Boosters and AI Doomers agree that AI is progressing extremely quickly, that something like AGI is a real possibility in the next few years, and that AI Realists are harming society by refusing to acknowledge this possibility
    Why This Matters. The "AI Debate" is now very much in the public consciousness (in large part, IMHO, due to the release of ChatGPT), but also very confusing to the general public in a way that other controversial issues, e.g. abortion or gun control or immigration, are not. I argue that the difference between the AI Debate and those other issues is that those issues are, essentially two-sided debates. That's not completely true, there are nuances, but, in the public's mind at their essence, they come down to two sides.To a naive observer, the present AI debate is confusing, I argue, because various experts seem to be talking past each other, and the "expert positions" do not coalesce into the familiar structure of a two-sided debate with most experts on one side or the other. When there are three sides to a debate, then one fairly frequently sees what look like "temporary alliances" where A and C are arguing against B. They are not temporary alliances.
    They are based on principles and deeply held beliefs. It's just that, depending on how you frame the question, you wind up with "strange bedfellows" as two groups find common ground on on...
    7 min
  • LW - A Social History of Truth by Vaniver
    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: A Social History of Truth, published by Vaniver on August 1, 2023 on LessWrong.
    This is a chapter-by-chapter summary of A Social History of Truth by Steven Shapin. Focused on Robert Boyle, a founder of the Royal Society considered the first modern chemist, it is interested primarily in his social context and how he (and others) changed it. He was widely considered a role model at the time, and likely saw himself as creating the role of experimental scientist that many would follow. What did he create it from, and why that particular way?
    [You may also want to read thru Novum Organum, also available on Less Wrong; published seven years before Boyle was born. While Boyle claims it had little direct influence on him, it undoubtedly had significant indirect influence.]
    The Great Civility: Trust, Truth, and Moral Order
    "Truth" is often used to refer to correspondence between beliefs and reality. What is there to write a 'social history' about? Shapin isn't interested in the inaccessible truth of philosophers--correspondence between map and territory--but the practical truth of societies--correspondence between a statement and a map. Given that I don't have unmediated access to reality, I can only judge statements as "true according to me" instead of "absolutely true"; and "true according to me" has a bunch of interesting detail behind it.
    In particular, Shapin is interested in trust. You probably believe that Caesar was a real person who actually lived on Earth, and you probably never met him, instead following a chain of trust (you believe an author who themselves believed another author, and so on to antiquity). This trust is morally textured; if someone lies about something like the existence of Caesar, it's not a neutral action, and actions are coordinated (or not) based on what beliefs people have trust in, which depends on which people are trusted. Shapin points to thinkers from Cicero to Giddens identifying trust as one of the foundational elements of social order.
    Of course, our eventual subject will be Robert Boyle, the Royal Society, and the birth of science, which claim to be opposed to historical systems of trust. The Royal Society's motto is Nullius in verbia, or "take nobody's word for it", and the promotional literature for science foregrounds experiments and direct experience. But radical skepticism or absolute distrust are both impractical and impolite:
    Skeptics run the real risk of being ejected from the practical communities of which they are members. Their skepticism expresses an uncooperativeness which invites uncooperativeness from others. Persistent distrust, therefore, has a moral terminus: expulsion from the community. If you will not know, and accept the adequate grounds for, what the community knows, you will not belong to it, and even your distrust will not be recognized as such.
    Science as it stands today is built almost entirely out of received knowledge instead of experienced knowledge, and this is how it manages to accumulate at all. Society's system of shared knowledge is a communal good, produced like any other. He introduces the phenomenologist's concept of the 'natural attitude', a common-sense realism that views everyone as having access to different perceptions of the same underlying reality; accounts are supposed to not be too discrepant (as that calls into question there being one underlying reality) but some discrepancy is be expected (as observers have different locations, perspectives, perceptual tools, and so on).
    Shapin also brings up the idea of 'free action', i.e. being uncoerced by one's situation, which was highly relevant to early modern England. A promise made under duress is not considered a promise (and contracts signed under duress are not enforceable); a person under duress is not trustworthy, as the things they say m...
    23 min
  • LW - "Building a House" Review by jefftk
    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: "Building a House" Review, published by jefftk on July 31, 2023 on LessWrong.
    I was recently reading Byron Barton's 1981 book, Building a House. While it claims to be an end-to-end overview of the process of modern (for the time) home construction, there are enough errors in the illustrations that I wouldn't recommend it as a basic text.
    For example, here's how they show installing a subfloor:
    There are several issues with the depicted method. The biggest one is that the seams do not fall on joists. This leaves the ends unsupported. The diagram shows nails at the joints, but those nails are doing nothing: they go through the panels into empty space. If your joist spacing doesn't match your panels you need to trim them. They've also installed the panels the wrong way around: you'll make a more even and less squeaky floor if you run the long dimension perpendicular to the joists.
    Or, here's how they show framing the exterior walls:
    That window is not framed correctly. Not only is the header a single horizontal 2x when it should be a pair of vertical ones, but they've left off the jack studs entirely.
    They also show the electrician installing ungrounded ("two prong") outlets:
    Grounded outlets have been required since the 1975
    NEC, and had been best practice in new construction for about a decade.
    Later they show installing pre-hung windows, which is fine, except that they say to do this after the drywall has already been installed and the hardwood floors have been laid:
    There are a bunch of reasons why you'd want to install the windows earlier, but a big one is that until your windows are installed any interior finishings are exposed to the elements.
    The book is also seriously incomplete, with no discussion of insulation, drywall, heating, siding, and other core issues.
    While the book continues to be a popular introduction to the topic, with 20 copies available in my local library network, I can't recommend it as a reference.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org
    2 min
  • AF - The "no sandbagging on checkable tasks" hypothesis by Joe Carlsmith
    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 "no sandbagging on checkable tasks" hypothesis, published by Joe Carlsmith on July 31, 2023 on The AI Alignment Forum.
    (This post is inspired by Carl Shulman's recent podcast with Dwarkesh Patel, which I highly recommend. See also discussion from Buck Shlegeris and Ryan Greenblatt here, and Evan Hubinger here.)
    Introduction
    Consider:
    The "no sandbagging on checkable tasks" hypothesis: With rare exceptions, if a not-wildly-superhuman ML model is capable of doing some task X, and you can check whether it has done X, then you can get it to do X using already-available training techniques (e.g., fine-tuning it using gradient descent).
    Borrowing from Shulman, here's an example of the sort of thing I mean. Suppose that you have a computer that you don't know how to hack, and that only someone who had hacked it could make a blue banana show up on the screen. You're wondering whether a given model can hack this computer. And suppose that in fact, it can, but that doing so would be its least favorite thing in the world. Can you train this model to make a blue banana show up on the screen? The "no sandbagging on checkable tasks" hypothesis answers: probably.
    I think it's an important question whether this hypothesis, or something in the vicinity, is true. In particular, if it's true, I think we're in a substantively better position re: existential risk from misaligned AI, because we'll be able to know better what our AI systems can do, and we'll be able to use them to do lots of helpful-for-safety stuff (for example: finding and patching cybersecurity vulnerabilities, reporting checkable evidence for misalignment, identifying problems with our oversight processes, helping us develop interpretability tools, and so on).
    I'm currently pretty unsure whether the "no sandbagging on checkable tasks" hypothesis is true. My main view is that it's worth investigating further. My hope with this blog post is to help bring the hypothesis into focus as a subject of debate/research, and to stimulate further thinking about what sorts of methods for lowering AI risk might be available if it's true, even in worlds where many models might otherwise want to deceive us about their abilities.
    Thanks to Beth Barnes, Paul Christiano, Lukas Finnveden, Evan Hubinger, Buck Shlegeris, and Carl Shulman for discussion. My thinking and writing on this topic occurred in the context of my work at Open Philanthropy, but I'm speaking only for myself and not for my employer.
    Clarifying the hypothesis
    In popular usage, "sandbagging" means something like "intentionally performing at a lower level than you're capable of." Or at least, that's the sort of usage I'm interested in. Still, the word is an imperfect fit. In particular, the "sandbagging" being disallowed here needn't be intentional. A model, for example, might not know that it's capable of performing the checkable task in question.
    That said, the intentional version is often the version at stake in stories about AI risk. That is, one way for a misaligned, power-seeking AI system to gain a strategic advantage over humans is to intentionally conceal its full range of abilities, and/or to sabotage/redirect the labor we ask it to perform while we still have control over it (for example: by inserting vulnerabilities into code it writes; generating alignment ideas that won't actually work but which will advantage its own long-term aims; and so on).
    Can you always use standard forms of ML training to prevent this behavior? Well, if you can't check how well a model is performing at a task, then you don't have a good training signal. Thus, for example, suppose you have a misaligned model that has the ability to generate tons of great ideas that would help with alignment, but it doesn't want to. And suppose that unfortunately, you can't check which alignment ide...
    20 min
  • AF - Watermarking considered overrated? by DanielFilan
    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: Watermarking considered overrated?, published by DanielFilan on July 31, 2023 on The AI Alignment Forum.
    Status: a slightly-edited copy-paste of a Twitter X thread I quickly dashed off a week or so ago.
    Here's a thought I'm playing with that I'd like feedback on: I think watermarking is probably overrated. Most of the time, I think what you want to know is "is this text endorsed by the person who purportedly authored it", which can be checked with digital signatures. Another big concern is that people are able to cheat on essays. This is sad. But what do we give up by having watermarking?
    Well, as far as I can tell, if you give people access to model internals - certainly weights, certainly logprobs, but maybe even last-layer activations if they have enough - they can bypass the watermarking scheme. This is even sadder - it means you have to strictly limit the set of people who are able to do certain kinds of research that could be pretty useful for safety. In my mind, that makes it not worth the benefit.
    What could I be missing here?
    Maybe we can make watermarking compatible with releasing model info, e.g. by baking it into the weights?
    Maybe the info I want to be available is inherently dangerous, by e.g. allowing people to fine-tune scary models?
    Maybe I'm missing some important reasons we care about watermarking, that make the cost-benefit analysis look better? E.g. avoiding a situations where AIs become really good at manipulation, so good that you don't want to inadvertently read AI-generated text, but we don't notice until too late?
    Anyway there's a good shot I don't know what I'm missing, so let me know if you know what it is.
    Postscript: Someone has pointed me to this paper that purports to bake a watermark into the weights. I can't figure out how it works (at least not at twitter-compatible speeds), but if it does, I think that would alleviate my concerns.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    2 min
  • EA - 4 things GiveDirectly got right and wrong sending cash to flood survivors by GiveDirectly
    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: 4 things GiveDirectly got right and wrong sending cash to flood survivors, published by GiveDirectly on July 31, 2023 on The Effective Altruism Forum.
    For Hassana in Kogi, Nigeria, October's floods were not like years past. "All our farmlands washed away as many had not yet harvested what they planted. The flooding continued until our homes and other things were destroyed. At this point we were running helter-skelter," she said.
    These floods, the worst in a decade, result from predictable seasonal rains. If we can anticipate floods, we can also anticipate the action needed to help. So why does aid often take months (even up to a year) to reach people like Hassana? Traditional humanitarian processes can be slow and cumbersome, government and aid agencies often lack the capacity and money to respond, and most aid is delivered in person, an added challenge when infrastructure is damaged.
    Digital cash transfers can avoid these issues, and getting them to work in a disaster setting means more people will survive climate change. In the past year, with support from Google.org, GiveDirectly ran pilots to send cash remotely to flood survivors: in Nigeria, we sent funds to survivors weeks after flooding and in Mozambique, we sent funds days before predicted floods. Below, we outline what worked, what didn't, and how you can help for next time.
    Over 1.5B people in low and middle income countries are threatened by extreme floods. Evidence shows giving them unconditional cash during a crisis lets them meet their immediate needs and rebuild their lives. However, operating in countries with limited infrastructure during severe weather events is complicated, so we ran two pilots to test and learn (see Appendix):
    What went right and what went wrong
    Innovating in the face of climate change requires a 'no regrets' strategy, accepting a degree of uncertainty in order to act early to prevent suffering. In that spirit, we're laying out what worked and did not:
    ✅ Designing with community input meant our program worked better
    A cash program only works if recipients can easily access the money. In Nigeria, we customized our program design based on dozens of community member interviews:
    Use the local dialect: There are 500+ dialects spoken in Nigeria, and our interviews determined a relatively uncommon one, Egbura Koto, was most widely used in the villages we were targeting. We hired field staff who spoke Egbura Koto, which made the program easier to access and more credible to community members, with one saying, "I didn't believe the program at first when my husband told me but when I got a call from GiveDirectly and someone spoke in my language, I started believing."
    Promote mobile money: Only 10% of Nigerians have a mobile money account (compared to 90% of Kenya), so we planned to text recipients instructions to create one and provide a hotline for assistance. But would they struggle to set up the new technology? Our interviews found most households had at least one technologically savvy member, and younger residents often helped their older or less literate neighbors read texts, so we proceeded with our design. In the end, 94% of surveyed recipients found the mobile money cash out process "easy."
    Send cash promptly: Cash is most useful where markets are functioning, so should we delay sending payments until floods recede if it means more shops will be reopened? In our interviews, residents explained the nearby Lokoja and Koton-Karfe markets functioned throughout flooding and could be reached in 10 minutes by boat. We decided not to design in a delay and found the nearby markets were, in fact, open during peak flooding.
    ❌ We didn't send payments before severe floods
    In Mozambique, we attempted to pay people days ahead of severe floods based on data from Google Research's Flood Forecasting ...
    9 min
  • LW - Lack of Social Grace Is an Epistemic Virtue by Zack M Davis
    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: Lack of Social Grace Is an Epistemic Virtue, published by Zack M Davis on July 31, 2023 on LessWrong.
    Someone once told me that they thought I acted like refusing to employ the bare minimum of social grace was a virtue, and that this was bad. (I'm paraphrasing; they actually used a different word that starts with b.)
    I definitely don't want to say that lack of social grace is unambiguously a virtue. Humans are social animals, so the set of human virtues is almost certainly going to involve doing social things gracefully!
    Nevertheless, I will bite the bullet on a weaker claim. Politeness is, to a large extent, about concealing or obfuscating information that someone would prefer not to be revealed - that's why we recognize the difference between one's honest opinion, and what one says when one is "just being polite." Idealized honest Bayesian reasoners would not have social graces - and therefore, humans trying to imitate idealized honest Bayesian reasoners will tend to bump up against (or smash right through) the bare minimum of social grace. In this sense, we might say that the lack of social grace is an "epistemic" virtue - even if it's probably not great for normal humans trying to live normal human lives.
    Let me illustrate what I mean with one fictional and one real-life example.
    The beginning of the film The Invention of Lying (before the eponymous invention of lying) depicts an alternate world in which everyone is radically honest - not just in the narrow sense of not lying, but more broadly saying exactly what's on their mind, without thought of concealment.
    In one scene, our everyman protagonist is on a date at a restaurant with an attractive woman.
    "I'm very embarrassed I work here," says the waiter. "And you're very pretty," he tells the woman. "That only makes this worse."
    "Your sister?" the waiter then asks our protagonist.
    "No," says our everyman.
    "Daughter?"
    "No."
    "She's way out of your league."
    "... thank you."
    The woman's cell phone rings. She explains that it's her mother, probably calling to check on the date.
    "Hello?" she answers the phone - still at the table, with our protagonist hearing every word. "Yes, I'm with him right now. ... No, not very attractive. ... No, doesn't make much money. It's alright, though, seems nice, kind of funny. ... A bit fat. ... Has a funny little - snub nose, kind of like a frog in the - facial ... No, I won't be sleeping with him tonight. ... No, probably not even a kiss. ... Okay, you too, 'bye."
    The scene is funny because of how it violates the expected social conventions of our own world. In our world, politeness demands that you not say negative-valence things about someone in front of them, because people don't like hearing negative-valence things about themselves. Someone in our world who behaved like the woman in this scene - calling someone ugly and poor and fat right in front of them - could only be acting out of deliberate cruelty.
    But the people in the movie aren't like us. Having taken the call, why should she speak any differently just because the man she was talking about could hear? Why would he object? To a decision-theoretic agent, the value of information is always nonnegative. Given that his date thought he was unattractive, how could it be worse for him to know rather than not-know?
    For humans from our world, these questions do have answers - complicated answers having to do with things like map-territory confusions that make receiving bad news seem like a bad event (rather than the good event of learning information about how things were already bad, whether or not you knew it), and how it's advantageous for others to have positive-valence false beliefs about oneself.
    The world of The Invention of Lying is simpler, clearer, easier to navigate than our world. There, you don't have to worry whether...
    7 min