The Nonlinear Library

The Nonlinear Library

By The Nonlinear FundEducation
Download on the App Store

All The Nonlinear Library episodes

  • LW - Clarifying Your Principles 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: Clarifying Your Principles, published by Raemon on October 1, 2022 on LessWrong.
    Following up on:
    Integrity and Accountability are Core Parts of Rationality
    Robust Agency for People and Organizations
    It's easy to end up in situations where people don't know which principles you stand for. (And, which ones you'll really stand for when it's inconvenient).
    It's easy to end up in situations where you don't know what principles you stand for, and which ones you'll actually stand up for when it's inconvenient.
    You can get away without knowing exactly what your principles are, for awhile. But if you want others to trust you in high stakes situations, it's helpful if you've proactively made your principles clear, and demonstrated that you can live by them. Most people don't really stand up for principles when it's inconvenient, so your prior should be that you probably won't, either.
    Being clear about your principles requires you to have them in the first place, and to practice the muscle of standing up for them, so that you know they are real and not just a vague applause light you are professing.
    This post is downstream of ideas I gained from Andrew Critch, Duncan Sabien, and Oliver Habryka, and benefitted from a lot of discussion with Elizabeth Van Nostrand (Though none of them necessarily endorse this essay). It is also somewhat inspired by Eliezer's Meta-Honesty: Firming Up Honesty Around Its Edge-Cases.
    Integrity Debt
    In software development, sometimes you need to write messy code that will cause problems for you later on down the line. Writing "good code" would take too long, and you need to ship your product. But, eventually, this messiness is going to make it harder for you to make progress on the codebase, and you'll want to spend some time simplifying your code to "pay down the debt."
    Similarly, if you're starting a new organization that depends on trust (either with the public, or particular stakeholders), there's a bunch of actions you might take to build that trust... which you may not have time to do when you're getting started.
    Integrity debt accumulates most acutely when you make controversial judgment calls that are hard to explain. Or, judgments that you reflectively wouldn't endorse. I think it also accumulates in small ways, when you do non-controversial but nonetheless confusing things.
    But perhaps more importantly (if subtly), integrity debt accumulates when you take on responsibilities that require you to have principles, without yet knowing what those principles are. This may work initially, but eventually will be like building a castle on a foundation of sand. Sooner or later you need to figure out the principles underlying you, your project, or your organization.
    If you've undertaken responsibilities without understanding the principles that will guide you in edge-cases, you may find that people have lost trust in you. Or, you may find you have lost trust in yourself.
    You may need to pay down your integrity debt. Or, alternately – declare bankruptcy, and transition into strategies that don't depend on trust. Which of these is the right strategy depends on your situation.
    Disclaimers:
    1. There is a moral element to this, but a lot of my motivation here is figuring out "how do we improve coordination in a low-trust world?". People can disagree on what is morally commendable, but improve their ability to coordinate on net-improvements, even with people they disapprove of.
    2. This article jumps back and forth between talking about Integrity, Trust, Accountability and Transparency. These are all different things, and I think you can have each one without the others. But I think they naturally fit together in particular ways that form some obvious strategies.
    I'm using 'integrity' and 'accountability' in ways similar to habryka. A quick recap:
    When I say...
    15 min
  • AF - Paper: Large Language Models Can Self-improve [Linkpost] by Evan R. Murphy
    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: Paper: Large Language Models Can Self-improve [Linkpost], published by Evan R. Murphy on October 2, 2022 on The AI Alignment Forum.
    This paper came out about a week ago. I am not the author - it was published anonymously.
    I learned about the paper when JJ Hepburn shared it on Slack. I just thought it seemed potentially really important and I hadn't seen it discussed on this forum yet:
    Paper title: "Large Language Models Can Self-improve"
    Author: Anonymous
    Abstract: "Large Language Models (LLMs) have achieved excellent performances in various tasks. However, fine-tuning an LLM requires extensive supervision. Human, on the other hand, may improve their reasoning abilities by self-thinking without external inputs. In this work, we demonstrate that an LLM is also capable of self-improving with only unlabeled datasets. We use a pre-trained LLM to generate “high-confidence” rationale-augmented answers for unlabeled questions using Chain-of-Thought prompting and self-consistency, and fine-tune the LLM using those self-generated solutions as target outputs. We show that our approach improves the general reasoning ability of a 540B-parameter LLM (74.4%82.1% on GSM8K, 78.2%83.0% on DROP, 90.0%94.4% on OpenBookQA, and 63.4%67.9% on ANLI-A3) and achieves state-of-the-art-level performance, without any ground truth label. We conduct ablation studies and show that finetuning on reasoning is critical for self-improvement."
    Link:
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    2 min
  • EA - EA is Not Religious Enough (EA should emulate peak Quakerism) by Lawrence Newport
    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: EA is Not Religious Enough (EA should emulate peak Quakerism), published by Lawrence Newport on October 1, 2022 on The Effective Altruism Forum.
    Here is a post from my blog which I'd really appreciate community thoughts on - and what it might look like if we were to seriously examine how to repeat some of the policy victories of past movements of this kind. Essentially, I believe that using examples from the past can provide us with blueprints for effective actions. This is especially true the more that EA wants to grow, and the more that longtermism implies both policy actions and longevity of the movement. Text is below (I've deleted the footnotes for slightly more readability on the forums):
    Pretty much every criticism of Effective Altruism has some claim that EA “is a lot like a religion”. This is a strange and useless criticism. Religion is just an example of what it looks like when people believe things. If you watch political people (or even sometimes sports fans) together in large groups, multiple aspects of their behaviour look decidedly religious - they address one another with special terms, they have chants, sing songs, have sacred texts or even sacred unimpeachable characters, they adorn themselves in signals of their beliefs, they pontificate about perfect worlds in which their policy (or victory) is fully realised and everyone lives happily ever after, and they have special beliefs built upon assumptions they protect emotionally regardless (frequently) of evidence - it doesn’t mean these are religions. However, I think EA does have a religion problem, namely, that it isn’t religious enough.
    Tyler Cowen argued at EAG DC that EAs should “be more Mormon”. His point was, essentially, that the historical lessons from Mormon communities showed clear upsides and thrust people in the general direction of “doing good”. In other words, the vibe of Mormonism was probably an overall good vibe and we shouldn’t spend forever analysing a priori what is best to do - we can learn from history and emulate successful groups. (I should note, he didn’t use the word vibe, but that is a good description of the approach to take to large historical trends we want to emulate). I agree with his criticism that EAs should weigh, understand and emulate suitable historical precedents and learn more clearly from them, but I strongly disagree that Mormonism, or another established church, is the right fit. Instead: EAs should emulate Quakerism. There are very good reasons for EAs to consider this history as the closest set of lessons they could learn from.
    Doing Longterm Good as a Minority View
    Quakerism, since its inception, was a reforming and productive presence. Quakerism, at its heart, is essentially a radical form of protestantism. Quakers believe that no individual has a privileged access to the word of God and thus everyone is to be listened to, heard and able to speak. It is, in many ways, more radical than evangelicals today - it takes further the concept of individual interpretation of the bible and instead moves this to individual interpretation writ-large. God speaks through everyone, and it is only through reflection, consideration and understanding of multiple views, that the Quaker community believes it can better understand what it is to do.
    This belief has had outsized positive effects since its inception. George Fox proposed these ideas in the English civil wars, a time period wherein radical protestants were voicing increasingly violent and concerning things. Inspired by George Fox, the Quakers arise, continue that radical protestant logic a step further, emphasising how this actually leads to a belief in toleration of views and non-violence and set about trying to convince essentially ISIS-level protestant groups that “hey, maybe you should be non-violent and actu...
    15 min
  • EA - Strange Love - Developing Empathy With Intention (or: How I Learned To Stop Calculating And Love The Cause) by EdoArad
    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: Strange Love - Developing Empathy With Intention (or: How I Learned To Stop Calculating And Love The Cause), published by EdoArad on October 1, 2022 on The Effective Altruism Forum.
    Tl;dr: if you are intellectually convinced that a cause is important, but you don’t emotionally empathize with the people or animals affected, you can develop this emotional empathy intentionally.
    When Andrés Jiménez Zorrilla first heard that Charity Entrepreneurship wanted to incubate a charity devoted to shrimp welfare, he was pretty sceptical. He’d cared about animal rights for a long time, but this still seemed like a deeply weird idea. However, over the course of the CE incubation program, he learnt some things which shifted his perspective. He discovered that 350 billion shrimp per year are farmed for food. Even if there was only a tiny chance that shrimp are sentient, astronomical numbers like that meant that he should take the issue seriously.
    Andrés gradually became intellectually convinced that shrimp welfare could be a very important cause, but he was not yet emotionally invested. He decided that it was important to emotionally connect to the cause. When covid lockdowns lifted, he went to visit a shrimp farm and actually witness the beings he would be helping, an experience that he found extremely powerful.
    He now runs the Shrimp Welfare Project. He says that he feels deep empathy for shrimp, as well as being intellectually convinced that it is important to improve their welfare. He even has a shrimp tattoo on his arm!
    EAs sometimes say that they feel intellectually compelled by the arguments that they should care about the suffering of non-human animals, future people, digital beings, or some other group of sentients who are outside of our typical moral circle, but they struggle to have emotional empathy for the suffering of this group.
    However, in my experience, this is not a fixed fact about a person - you can work to develop empathy for beings who are very different to you, or distant from you in time or space. In this post, we offer some advice on how to do this.
    Why should I develop more empathy?
    It’s not necessarily a problem to have a mismatch between your beliefs and your emotions.
    In Radical Empathy, Holden Karnofsky suggests that we should care intellectually about many beings and causes that we don’t care about emotionally. The opposite is also true - in his book Against Empathy, Paul Bloom argues that our intuitive emotional caring might lead to bad choices. EAs are generally somewhat suspicious of ‘warm fuzzies’, and it’s a central aspect of EA that we should be guided more by our brains than our guts when it comes to altruism.
    However, there are some reasons why people might want their empathy to match their beliefs:
    Congruence: it’s easier to decide what to do if our emotional intuitions and our considered beliefs line up. When they conflict, we are more unsure about how to act, or how to think about unfamiliar scenarios.
    Motivation: you might feel unmotivated to work on the causes you care about if you don’t feel enough empathy for beings who benefit
    Understanding other EAs: even if you don’t want to work on (for example) ending insect suffering yourself, developing some empathy for insects might help you better understand people who do
    This post is about aligning our intuitive emotional empathy with our cognitive, intellectual empathy.
    How can I develop more empathy?
    Here are some suggestions for how to develop empathy. In general, detailed knowledge creates empathy (particularly for people who are already empathetic by nature). If you want to empathize with a certain type of being, try to find out more about their lives and their troubles, or imagine in detail what it’s like to be them.
    Learn more about the beings you want to empathize with 🦐
    When I (Edo) hea...
    8 min
  • AF - Four usages of "loss" in AI by Alex Turner
    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: Four usages of "loss" in AI, published by Alex Turner on October 2, 2022 on The AI Alignment Forum.
    Summary: What does it mean for a loss function to be "aligned with" human goals? I perceive four different concepts which involve "loss function" in importantly different ways:
    Physical-loss: The physical implementation of a loss function and the loss computations,
    Mathematical-loss: The mathematical idealization of a loss function,
    A loss function "encoding/representing/aligning with" an intended goal, and
    Agents which "care about achieving low loss."
    I advocate retaining physical- and mathematical-loss. I advocate dropping 3 in favor of talking directly about desired AI cognition and how the loss function entrains that cognition. I advocate disambiguating 4, because it can refer to a range of physically grounded preferences about loss (e.g. low value at the loss register versus making perfect future predictions).
    Related: Towards deconfusing wireheading and reward maximization. I'm going to talk about "loss" instead of "reward", but the lessons apply to both.
    I think it's important to maintain a sharp distinction between the following four concepts.
    1: Physically implemented loss
    The loss function updated my network.
    This is a statement about computations embedded in physical reality. This statement involves the physically implemented sequence of loss computations which stream in throughout training. For example, the computations engendered by loss_fn = torch.nn.CrossEntropyLoss().
    2: Mathematical loss
    The loss function is a smooth function of the prediction distribution.
    This is a statement about the idealized mathematical loss function. These are the mathematical objects you can prove learning theory results about. The Platonic idealization of the learning problem and the mathematical output-grading rule casts a shadow into your computer via its real-world implementation (concept 1).
    For example, (D,ℓ) where D:={(x,label(x))∣x∈MNIST} is the mathematical idealization of the MNIST dataset, where the x∈R28×28 are the idealized grayscale MNIST images. And ℓ is the mathematical function of cross-entropy (CE) loss between a label prediction distribution and the ground-truth labels.
    3: Loss functions "representing" goals
    I want a loss function which is aligned with the goal of "write good novels."
    This is an aspirational statement about achieving some kind of correspondence between the loss function and the goal of writing good novels. But what does this statement mean?
    Suppose you tell me "I have written down a loss function ℓnovel which is perfectly aligned with the goal of 'write good novels'." What experiences should this claim lead me to anticipate?
    That an agent can only achieve low physical-loss (concept 1) if it has, in physical fact, written a good novel?
    That in some mathematical idealization of the learning problem (concept 2), loss-minimization only occurs when the agent outputs text which would be found in what we rightly consider to be "good novels"? (But in which mathematical idealization?)
    That, as a matter of physical fact, if you train an agent on ℓnovel using learning setup X, then you produce a language model which can be easily prompted to output high-quality novels?
    The imprecision comes from loss functions not directly encoding goals. Loss signals are physically implemented (concept 1) parts of the AI's training process which (physically) update the AI's cognition in certain ways. While a loss function can be involved in the AI's decision-making structure (see items i and ii above), additional information is needed to understand what motivational cognition is being discussed.
    I think that talking about loss functions being "aligned" encourages bad habits of thought at best, and is nonsensical at worst. I think it makes way more sense to say how you ...
    10 min
  • LW - What sorts of preparations ought I do in case of further escalation in Ukraine? by tailcalled
    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: What sorts of preparations ought I do in case of further escalation in Ukraine?, published by tailcalled on October 1, 2022 on LessWrong.
    I guess the Russian invasion of Ukraine can go in many ways. It might just fizzle out, or continue to be a fairly conventional war with extent limited to Ukraine, or it might escalate with a Russian nuclear strike against Ukraine followed by NATO counterattacks against Russia, and it doesn't seem implausible to me that it could spin totally out of control into a WW3 scenario. Or various other paths.
    The more extreme end of the range with these scenarios seem worth preparing for, but I'm not completely sure what preparations would be the most relevant. I thought LessWrong might have some good ideas.
    If relevant, I live in Denmark, a NATO country close to the core of the EU, not quite at the border of Russia but also not quite buffered away from Russia (especially if counting Kaliningrad). Though preparation ideas that are more relevant to people in other countries than Denmark are welcome too as they might be useful to someone else.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    2 min
  • EA - EA may look like a cult (and it's not just optics) by nadavb
    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: EA may look like a cult (and it's not just optics), published by nadavb on October 1, 2022 on The Effective Altruism Forum.
    Is EA a cult?
    Cultishness is a spectrum. As I’ll demonstrate, EA has some characteristics normally associated with cults, which can give both outsiders and those already engaged with EA a very bad impression. I'm not arguing though that the situation is so bad that EA is a cult in the extreme sense of the word. But I think that people who don’t know much about EA and get the impression that it’s a cult are not crazy to think that.
    A narrative of EA as a cult
    What if I told you about a group of people calling themselves Effective Altruists (EAs) who wish to spread their movement to every corner of the world. Members of this movement spend a lot of money and energy on converting more people into their ideology (they call it community building). They tend to target young people, especially university students (and increasingly high school students). They give away free books written by their gurus and run fellowship programs. And they will gladly pay all the expenses for newcomers to travel and attend their events so that they can expose them to their ideas.
    EAs are encouraged to consult the doctrine and other movement members about most major life decisions, including what to do for a living (career consultation), how to spend their money (donation advice), and what to do with their spare time (volunteering). After joining the movement, EAs are encouraged to give away 10% or more of their income to support the movement and its projects. It’s not uncommon for EAs to socialize and live mostly with other EAs (some will get subsidized EA housing). Some EAs want even their romantic partners to be members of the community.
    While they tend to dismiss what’s considered common sense by normal mainstream society, EAs will easily embrace very weird-sounding ideas once endorsed by the movement and its leaders.
    Many EAs believe that the world as we know it may soon come to an end and that humanity is under existential threat. They believe that most normal people are totally blind to the danger. EAs, on the other hand, have a special role in preventing the apocalypse, and only through incredible efforts can the world be saved. Many of these EAs describe their aspirations in utopian terms, declaring that an ideal world free of aging and death is waiting for us if we take the right actions. To save the world and navigate the future of humanity, EAs often talk about the need to influence governments and public opinion to match their beliefs.
    It’s not just optics
    While I’ve focused on how EA might be perceived from the outside, I think that many of the cult-like features of EA pose a real issue, so it’s not just a PR problem. I don’t yet have a good mental model of all the ways in which it plays out, but I believe there are other negative consequences. The impression of a cult could also explain why some of the recent media attention on EA hasn’t been very positive (I’m not saying it’s the only reason).
    How to make EA less cultish
    Many of the features of EA that make it look and sound like a cult (e.g. pursuit of growth, willingness to accept unconventional ideas) are quite essential to the project of doing the most good, so I’m not suggesting to automatically get rid of everything that could possibly be perceived as cultish (on the other hand, I think that how EA is perceived is quite important, so we shouldn’t ignore these considerations either). Having said that, I believe there are cultural norms we could embrace that would push us away from cultishness without significantly compromising other goals.
    Some helpful anti-clutishness norms that are already established in EA to some extent and I’d like to see further cultivated include:
    Critical thinking
    Diversity o...
    5 min
  • EA - What I learned from the criticism contest by Gavin
    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: What I learned from the criticism contest, published by Gavin on October 1, 2022 on The Effective Altruism Forum.
    I was a judge on the Criticism and Red-teaming Contest, and read 170 entries. It was overall great: hundreds of submissions and dozens of new points.
    Recurring patterns in the critiques
    But most people make the same points. Some of them have been made from the beginning, like 2011. You could take that as an indictment of EA's responsiveness to critics, proof that there's a problem, or merely as proof that critics don't read and that there's a small number of wide basins in criticism space. (We're launching the EA Bug Tracker to try and distinguish these scenarios, and to keep valid criticisms in sight.)
    Trends in submissions I saw:
    (I took out the examples because it was mean. Can back em up in DMs.)
    Academics are stuck in 2015. It's great that academics are writing full-blown papers about EA, and on average I expect this to help us fight groupthink and to bring new ideas in. But almost all of the papers submitted here are addressing a seriously outdated version of EA, before the longtermist shift, before the shift away from public calculation, before the systemic stuff. Some of them even just criticise Singer 2009 and assume this is equivalent to criticising EA. (I want to single out Sundaram et al as an exception. It is steeped in current EA while maintaining some very different worldviews.)
    Normalisation. For various reasons, many suggestions would make EA less distinctive. Whether that's intentional PR skulduggery, retconning a more mainstream cause into the tent, adding epicycles to show that mainstream problem x is really the biggest factor in AI risk, or just what happens when you average intuitions (the mode of a group will reflect current commonsense consensus about causes and interventions, and so not be very EA). This probably has some merit. But if we implemented all of these, we'd be destroyed.
    Schism. People were weirdly enthusiastic about schisming into two neartermist and longtermist movements. (They usually phrase this as a way of letting neartermist things get their due, but I see this as a sure way to doom it to normalisation instead.)
    Stop decoupling everything. The opposite mistake is to give up on decoupling, to allow the truism that 'all causes are connected' swamp focussed efforts.
    Names. People devote a huge amount of time to the connotations of different names. But obsessing over this stuff is an established EA foible.
    Vast amounts of ressentiment. Some critiques are just disagreements about cause prioritisation, phrased hotly as if this gave them more weight.
    EAs underestimate uncertainty in cause prioritisation. One perennial criticism which has always been true is that most of cause prioritisation, the heart of EA, is incredibly nonobvious and dependent on fiddly philosophical questions. And yet we don't much act like we knew this, aside from a few GPI economist-philosophers. This is probably the fairest criticism I hear from non-EAs.
    Fundamental criticism takes time
    Karnofsky, describing his former view: "Most EA criticism is - and should be - about the community as it exists today, rather than about the “core ideas.” The core ideas are just solid. Do the most good possible - should we really be arguing about that?" He changed his mind!
    Really fundamental challenges to your views don't move you at the time you read them. Instead they set dominoes falling; they alter some weights a little, so that the next time the problem comes up in your real life, you notice it and hold it in your attention for a fraction more of a second. And then over about 3 years, you become a different person, - and no trace of the original post remains, and no gratitude will accrue.
    If the winners of the contest don't strike you as fundamental critiqu...
    10 min
  • EA - Winners of the EA Criticism and Red Teaming Contest by Lizka
    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: Winners of the EA Criticism and Red Teaming Contest, published by Lizka on October 1, 2022 on The Effective Altruism Forum.
    We’re excited to announce the winners of the EA Criticism and Red Teaming Contest. We had 341 submissions and are awarding $120,000 in prizes to our top 31 entries.
    We set out with the primary goals of identifying errors in existing work in effective altruism, stress-testing important ideas, raising the average quality of criticism (in part to create examples for future work), and supporting a culture of openness and critical thinking. We’re pleased about the progress submissions to this contest made, though there’s certainly still lots of work to be done. We think the winners of the contest are both valuable in their own right as criticisms, and as helpful examples of different types of critique.
    We had a large judging panel. Not all panelists read every piece (even among the winners), and some pieces have won prizes despite being read by relatively few people or having some controversy over their value. Particularly when looking at challenges to the basic frameworks of effective altruism, there can be cases where there is significant uncertainty about whether a contribution is ultimately helpful. But if it is, it’s often very important, so we didn’t want to exclude cases like this from winning prizes when they had some strong advocates. You can read about our process and overall thoughts on the contest at the end of this post. Prize distribution logistics are also discussed at the end of this post.
    An overview of the winners
    Top prizes [see more]
    A critical review of GiveWell's 2022 cost-effectiveness model and Methods for improving uncertainty analysis in EA cost-effectiveness models by Alex Bates (Froolow) ($25,000 total)
    Biological Anchors external review by Jennifer Lin ($20,000)
    Population Ethics without Axiology: A Framework by Lukas Gloor ($20,000)
    Second prizes (runners up) — $5,000 each [see more]
    Are you really in a race? The Cautionary Tales of Szilárd and Ellsberg by Haydn Belfield
    Against Anthropic Shadow by Toby Crisford
    An Evaluation of Animal Charity Evaluators by eaanonymous1234
    Red Teaming CEA’s Community Building Work by AnonymousEAForumAccount
    A Critical Review of Open Philanthropy’s Bet On Criminal Justice Reform by Nuño Sempere
    Effective altruism in the garden of ends by Tyler Alterman
    Notes on effective altruism by Michael Nielsen
    Honorable mentions — $1,000 for each of the 20 in this category [see more]
    Top prizes
    A critical review of GiveWell's 2022 cost-effectiveness model and Methods for improving uncertainty analysis in EA cost-effectiveness models by Alex Bates (Froolow) ($25,000 prize in total)
    We’re awarding a total of $25,000 for these two submissions by the same author covering similar ground. A critical review of GiveWell’s 2022 cost-effectiveness model is a deep dive into the strengths and weaknesses of GiveWell’s analysis, and how it might be improved. Methods for improving uncertainty analysis in EA cost-effectiveness models extracts some more generalizable lessons.
    Summary of A critical review of GiveWell’s 2022 cost-effectiveness model: The submission replicates GiveWell’s cost-effectiveness models, critiques their design and structure, notes some minor errors, and suggests some broader takeaways for GiveWell and effective altruism. The author emphasizes GiveWell’s lack of uncertainty analysis as a weakness, notes issues with the models’ architectures (external data sources appear as inputs on many different levels of the model, elements from a given level in the model “grab” from others on that level, etc.), and discusses ways in which communication of the models is confusing. Overall, though, the author seems impressed with GiveWell’s work.
    You can also see the author’s own picture-based summary of thei...
    35 min
  • LW - A Few Terrifying Facts About The Russo-Ukrainian War by DonyChristie
    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 Few Terrifying Facts About The Russo-Ukrainian War, published by DonyChristie on September 30, 2022 on LessWrong.
    Epistemic status: trying to summarize the news and predict, post is under revision, too lazy to citation everything
    I wanted to collect a few observations I've made, as best I understand them. This PBS article does a good job of explaining much of it.
    Vladimir Putin has announced the annexation of four Ukrainian territories.
    This makes them Russian territory from Russia's perspective.
    “People living in Donetsk, Luhansk, Zaporizhzhia and Kherson are becoming our citizens forever” - Putin
    The West does not acknowledge this annexation, describing it as illegal. Ukraine does not acknowledge this annexation and says it plans to take the territories back.
    "By attempting to annex Ukraine's Donetsk, Luhansk, Zaporizhzhia and Kherson regions, (Russian President Vladimir) Putin tries to grab territories he doesn't even physically control on the ground. Nothing changes for Ukraine: we continue liberating our land and our people, restoring our territorial integrity," Ukraine's Foreign Minister Dmytro Kuleba said on social media.
    Russian military doctrine allows the usage of nuclear weapons to defend Russian territory.
    Putin has a track record of escalating apparently (this needs more data) and Russia seems to be planning for escalation until the war is won.
    "All of our sources in the elite — who all spoke on the condition of anonymity — said the military conflict will only escalate in the coming months."
    Putin has clearly stated that they will defend this territory, including with tactical nukes if need be.
    He said they would use "any means available" to defend it
    He has mentioned usage of nukes some number of times (a nice-to-have: a list of all the times he has said this)
    Medyedev has stated the West would not retaliate if nuclear weapons are used.
    "Under Russia’s amended constitution, no Kremlin leader can cede territories once they are annexed." - someone on Twitter
    Putin has stated he is not bluffing.
    The U.S. Secretary of State says it is "loose talk".
    Putin has called for a ceasefire.
    Ukraine and U.S. does not want to do this.
    The U.S. has said there will be "catastrophic consequences" if nuclear weapons are used.
    They are keeping the consequences vague for strategic flexibility.
    Concerning escalatory developments that aren't directly related to nuclear brinksmanship:
    The Nordstream natural gas pipes were blown up. We don't know who did it. (This section needs work)
    Russia could have done it
    Burning the bridges strategy?
    U.S. could have done it
    U.S. airships were nearby days before.
    Ukraine
    Some other country or group, hypothetically
    Russia has conscripted 300,000 men.
    There is some amount of resistance.
    Tens of thousands of people are leaving.
    There are some protests.
    Ukraine has "accelerated" its application to join NATO.
    Consensus from all 30 NATO countries is required, though.
    France and Germany have expressed reluctance in the past.
    "Experts warned that Ukraine’s NATO membership at the moment seems elusive at best. The process could take at least several months, and even years."
    Sweden and Finland have been approved by 28 out of 30 countries. Turkey and Hungary will likely hold out for a while.
    Any country in NATO that is attacked by Russia triggers the whole NATO alliance to attack Russia.
    Biden has affirmed this, naturally.
    Conclusion: Ukraine will keep attacking the annexed territories in order to take them back until Russia uses a tactical nuke out of desperation, and the U.S. will respond with "catastrophic consequences".
    This is obviously uncertain! But the chain of logic forms a coherent enough inside view for me to put a lot of probability on that, and I may start taking bets. I would be curious to see different inside views about what will happ...
    5 min