The Nonlinear Library

The Nonlinear Library

By The Nonlinear FundEducation
Download on the App Store

All The Nonlinear Library episodes

  • EA - [Linkpost] My moral view: Reducing suffering, 'how to be' as fundamental to morality, no positive value, cons of grand theory, and more - By Simon Knutsson by Alistair Webster
    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: [Linkpost] My moral view: Reducing suffering, 'how to be' as fundamental to morality, no positive value, cons of grand theory, and more - By Simon Knutsson, published by Alistair Webster on August 25, 2023 on The Effective Altruism Forum.
    The following are excerpts from a new essay by Simon Knutsson:
    Summary
    I explain parts of my moral view briefly. I also talk about some arguments for and against my view. The basics of my moral view include that one should focus on reducing severe suffering and behave well, and these basics get us pretty far in terms of how to act and be in real life.
    Introduction
    My moral view is suffering-focused in the sense that it emphasises the reduction of suffering and the like. My view might differ from other suffering-focused views in the following ways:
    I do not pick or try to formulate an overarching moral theory. Instead, my moral view is intentionally fairly non-theoretical.
    I think of ideas about how one should be, such as the idea that one should be considerate, as more primitive and fundamental to morality than some others seem to think of them. Many others agree that one should be like that, but perhaps for different reasons, such as that being like that has the best consequences.
    My take on what should be reduced for its own sake is perhaps unusually pluralistic - it is most important to reduce extreme suffering, but it also makes sense to reduce, for example, gruesome violence, ruined lives and life projects, and acts such as ignoring harms.
    My view could be labelled a pitch-black philosophical pessimism located towards the end of a philosophical optimism-pessimism spectrum. In my view, there is no positive value and no positive quality of life, and there are no positive experiences. An empty world is the best possible world, the world is terrible, and the future will almost certainly be appalling. (Of course, we should still try to make the world and the future less awful.)
    I am sceptical of categorical notions such as 'good' and 'positive value', and I instead prefer comparative notions such as 'better' and 'worse'.
    I don't think much in terms of uncertainty about moral principles or evaluative judgments; rather, I tend to think in terms of to what extent I accept or agree with specific ideas about morality and value.
    I think that the following are some of the advantages of my moral view: By being light on theory, it avoids pitfalls that high-level moral theories can have. It also takes suffering seriously, is overall reasonable, and lacks implausible implications such as when a view recommends that a clearly immoral act should be carried out. And my moral view is quite action-guiding in real life. This talk about advantages may sound like academic niceties, but most of the points have great practical importance. For example, it is crucial to direct one's attention and efforts to those who are or will be extremely badly off.
    Disadvantages of overarching moral theories
    I don't identify as a particularist or antitheorist, but I doubt the importance of overarching moral theories, and at least some of them seem to have drawbacks. That is, it seems we can do without high-level moral theories, and the following are three disadvantages that high-level moral theories can have (a caveat is that some high-level theories might be innocuous, and almost all examples of disadvantages I have observed concern consequentialist moral theories).
    The first disadvantage is when someone accepts a problematic implication of their favourite moral theory with the justification that "all moral theories have counterintuitive implications". The person holds that something is permissible even though one would normally think of it as immoral or even monstrous. But if all existing moral theories are so problematic, we need not choose any of them. Such accept...
    5 min
  • LW - AI #26: Fine Tuning Time by Zvi
    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: AI #26: Fine Tuning Time, published by Zvi on August 25, 2023 on LessWrong.
    GPT-3.5 fine tuning is here. GPT-4 fine tuning is only a few months away. It is about to get a lot easier to get a powerful system that does what you want it to do, and knows what you want it to know, especially for the purposes of a business or a website.
    As an experiment, I am putting in bold the sections I think are worth highlighting, as unusually important or interesting versions of the thing than in a typical week.
    Table of Contents
    Introduction.
    Table of Contents.
    Language Models Offer Mundane Utility. Claude-2 versus GPT-4.
    Language Models Don't Offer Mundane Utility. No opinions, no agents.
    Fact Check: Misleading. AI fact checker makes people more confused not less.
    GPT-4 Real This Time. Fine tune GPT-3.5, soon GPT-4. Ask it if it's sure.
    Fun With Image Generation. MidJourney inpainting ho. And oh no AI porn.
    Deepfaketown and Botpocalypse Soon. Adversarial examples starting to emerge.
    They Took Our Jobs. New York Times joins copyright lawsuits against OpenAI.
    Introducing. Palisade Research will study potentially dangerous AI affordances.
    In Other AI News. Who is adapting fastest to AI? An attempt to measure that.
    Quiet Speculations. Jack Clark asks questions about what the future will bring.
    The Quest for Sane Regulation. FTC asks OpenAI a different sort of question.
    The Week in Audio. It's win win.
    No One Would Be So Stupid As To. Make an AI conscious? Oh, come on.
    Aligning a Smarter Than Human Intelligence is Difficult. Evidence for IDA?
    People Are Worried About AI Killing Everyone. Polling numbers are very clear.
    The Lighter Side. Only half there.
    Language Models Offer Mundane Utility
    Which model is better, Claude-2 or GPT-4?
    Rowan Cheung makes the case that Claude 2 is superior. You get the 100k context window, ability to upload multiple files, data through early 2023 (versus late 2021) and faster processing time, all for free. In exchange, you give up plug-ins and it is worse at math. What Rowan does not mention is that GPT-4 has the edge in raw intelligence and general capability, and also the ability to set system instructions is helpful. He implies he isn't even paying the $20/month for GPT-4, which strikes me as insane.
    My verdict in practice is that by default I will use Claude-2. If I care about response quality I will use both and compare. When Claude-2 is clearly falling on its face, I'll go to GPT-4. On reflection, 'use both' is most often the correct strategy.
    He also looks at the plugins. There are so many plugins, at least 867 of them. Which are worth using?
    He recommends Zapier for automating through trigger actions, ChatWithPDF (I use Claude 2 for this), Wolfram Alpha for real-time data and math, VoxScript for YouTube video transcripts and web browsing, WebPilot which seems duplicative, Website Performance although I'm not sure why you'd use an AI for that, ScholarAI for searching papers, Shownotes to summarize podcasts (why?), ChatSpot for marketing and sales data and Expedia for vacation planning.
    I just booked a trip, and went on two others recently, and it didn't occur to me to use the Expedia plug-in rather than, among other websites, Expedia (my go-to plan is Orbitz for flights and Google Maps for hotels). Next time I should remember to try it.
    Study claims that salience of God increases acceptance of AI decisions. I would wait for the replication on this one. If it is true, it points out that there will be various ways for AIs to tip the scales towards us accepting their decisions, or potentially for humans to coordinate to turn against AI, that don't have much to do with any relevant considerations. Humans are rather buggy code.
    Matt Shumer recommends a GPT-4 system message.
    Use it to you help make engineering decisions in unfamiliar territory:
    You are an e...
    53 min
  • AF - A Model-based Approach to AI Existential Risk by Samuel Dylan Martin
    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 Model-based Approach to AI Existential Risk, published by Samuel Dylan Martin on August 25, 2023 on The AI Alignment Forum.
    Introduction
    Polarisation hampers cooperation and progress towards understanding whether future AI poses an existential risk to humanity and how to reduce the risks of catastrophic outcomes. It is exceptionally challenging to pin down what these risks are and what decisions are best. We believe that a model-based approach offers many advantages for improving our understanding of risks from AI, estimating the value of mitigation policies, and fostering communication between people on different sides of AI risk arguments. We also believe that a large percentage of practitioners in the AI safety and alignment communities have appropriate skill sets to successfully use model-based approaches.
    In this article, we will lead you through an example application of a model-based approach for the risk of an existential catastrophe from unaligned AI: a probabilistic model based on Carlsmith's Is Power-seeking AI an Existential Risk? You will interact with our model, explore your own assumptions, and (we hope) develop your own ideas for how this type of approach might be relevant in your own work. You can find a link to the model here.
    In many poorly understood areas, people gravitate to advocacy positions. We see this with AI risk, where it is common to see writers dismissively call someone an "AI doomer", or "AI accelerationist". People on each side of this debate are unable to communicate their ideas to the other side, since advocacy often includes biases and evidence interpreted within a framework not shared by the other side.
    In other domains, we have witnessed first-hand that model-based approaches are a constructive way to cut through advocacy like this. For example, by leveraging a model-based approach, the Rigs-to-Reefs project reached near consensus among 22 diverse organisations on the contentious problem of how to decommission the huge oil platforms off the Santa Barbara coast. For decades, environmental groups, oil companies, marine biologists, commercial and recreational fishermen, shipping interests, legal defence funds, the State of California, and federal agencies were stuck in an impasse on this issue. The introduction of a model refocused the dialog on specific assumptions, objectives and options, and led to 20 out of the 22 organisations agreeing on the same plan. The California legislature encoded this plan into law with bill AB 2503, which passed almost unanimously.
    There is a lot of uncertainty around existential risks from AI, and the stakes are extremely high. In situations like this, we advocate quantifying uncertainty explicitly using probability distributions. Sadly, this is not as common as it should be, even in domains where such techniques would be most useful.
    A recent paper on the risks of unaligned AI by Joe Carlsmith (2022) is a powerful illustration of how probabilistic methods can help assess whether advanced AI poses an existential risk to humanity. In this article, we review Carlsmith's argument and incorporate his problem decomposition into our own Analytica model. We then expand on this starting point in several ways to demonstrate elementary ways to approach each of the distinctive challenges in the x-risk domain. We take you on a tour of the live model to learn about its elements and enable you to dive deeper on your own.
    Challenges
    Predicting the long-term future is always challenging. The difficulty is amplified when there is no historical precedent. But this challenge is not unique; we lack historical precedent in many other areas, for example when considering a novel government program or a fundamentally new business initiative. We also lack precedent when world conditions change due to changes in technology, ...
    53 min
  • LW - AI Regulation May Be More Important Than AI Alignment For Existential Safety by otto.barten
    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: AI Regulation May Be More Important Than AI Alignment For Existential Safety, published by otto.barten on August 24, 2023 on LessWrong.
    Summary: Aligning a single powerful AI is not enough: we're only safe if no-one, ever, can build an unaligned powerful AI. Yudkowsky tried to solve this with the pivotal act: the first aligned AI does something (such as melting all GPUs) which makes sure no unaligned AIs can ever get built, by anyone. However, the labs are currently apparently not aiming to implement a pivotal act. That means that aligning an AGI, while creating lots of value, would not reduce existential risk. Instead, global hardware/data regulation is what's needed to reduce existential risk. Therefore, those aiming to reduce AI existential risk should focus on AI Regulation, rather than on AI Alignment.
    Epistemic status: I've been thinking about this for a few years, while working professionally on x-risk reduction. I think I know most literature on the topic. I have also discussed the topic with a fair number of experts (who in some cases seemed to agree, and in other cases did not seem to agree).
    Thanks to David Krueger, Matthijs Maas, Roman Yampolskiy, Tim Bakker, Ruben Dieleman, and Alex van der Meer for helpful conversations, comments, and/or feedback. These people do not necessarily share the views expressed in this post.
    This post is mostly about AI x-risk caused by a take-over. It may or may not be valid for other types of AI x-risks. This post is mostly about the 'end game' of AI existential risk, not about intermediate states.
    AI existential risk is an evolutionary problem. As Eliezer Yudkowsky and others have pointed out: even if there are safe AIs, those are irrelevant, since they will not prevent others from building dangerous AIs. Examples of safe AIs could be oracles or satisficers, insofar as it turns out to be possible to combine these AI types with high intelligence. But, as Yudkowsky would put it: "if all you need is an object that doesn't do dangerous things, you could try a sponge". Even if a limited AI would be a safe AI, it would not reduce AI existential risk. This is because at some point, someone would create an AI with an unbounded goal (create as many paperclips as possible, predict the next word in the sentence with unlimited accuracy, etc.). This is the AI that would kill us, not the safe one.
    This is the evolutionary nature of the AI existential risk problem. It is described excellently by Anthony Berglas in his underrated book, and more recently also in Ben Hendrycks' paper. This evolutionary part is a fundamental and very important property of AI existential risk and a large part of why this problem is difficult. Yet, many in AI Alignment and industry seem to focus on only aligning a single AI, which I think is insufficient.
    Yudkowsky aimed to solve this evolutionary problem (the fact that no-one, ever, should build an unsafe AI) with the so-called pivotal act. An aligned superintelligence would not only not kill humanity, it would also perform a pivotal act, the toy example being to melt all GPUs globally, or, as he later put it, to subtly change all GPUs globally so that they can no longer be used to create an AGI. This would be the act that would actually save humanity from extinction, by making sure no unsafe superintelligences are created, ever, by anyone (it may be argued that melting all GPUs, and all other future hardware that could run AI, would need to be done indefinitely by the aligned superintelligence, else even a pivotal act may be insufficient).
    The concept of a pivotal act, however, seems to have gone thoroughly out of fashion. None of the leading labs, AI governance think tanks, governments, etc. are talking or, apparently, thinking much about it. Rather, they seem to be thinking about things like non-proliferati...
    9 min
  • LW - The lost millennium by Ege Erdil
    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 lost millennium, published by Ege Erdil on August 24, 2023 on LessWrong.
    Note: This post is speculative, and you should take the claims in it with a grain of salt.
    Throughout human history, we know that economic growth, for which population growth was a good proxy under Malthusian conditions, has accelerated many times. Indeed, we can see this just by extrapolating current growth rates backward: given that there are only so many people in the world and they are only at a finite multiple of subsistence income per person, it would not be possible for economic growth rates on the order of 3% per year to have lasted for more than a thousand years or so.
    Though it's difficult to do so, people have also come up with estimates for exactly how this growth happened in the past. One such estimate is from the HYDE dataset, also available at Our World In Data; and a whole collection of other estimates can be found here.
    Let's examine the HYDE dataset first. The data before 1 AD is given per millennium, and looking at the growth rates gives us the table below:
    Millennium Mean annual population growth rate (percent) 10000 BC - 9000 BC 0.0236% 9000 BC - 8000 BC 0.0254% 8000 BC - 7000 BC 0.0279% 7000 BC - 6000 BC 0.0320% 6000 BC - 5000 BC 0.0368% 5000 BC - 4000 BC 0.0411% 4000 BC - 3000 BC 0.0435% 3000 BC - 2000 BC 0.0489% 2000 BC - 1000 BC 0.0415% 1000 BC - 1 AD 0.0746%
    Aside from a slight reversal in the second millennium BC[1], the estimates suggest a gradual acceleration in the growth rate of population. This would be consistent with e.g. a hyperbolic model for population growth, but here I don't want to make any such assumption and just look at the raw data.
    What is surprising, then, is the next entry that should be in this table: in the first millennium, 1 AD - 1000 AD, population growth is estimated to have averaged a mere 0.033%/year. In other words, according to this dataset, even the fifth millennium BC had faster population growth than the first millennium!
    Something like this result turns out to be robust to changing the dataset we're using. McEvedy and Jones report a mean growth rate of 0.044%/year for the first millennium, compared to 0.137%/year from 1000 BC to 200 BC and 0.062%/year from 200 BC to 1 AD. This more granular data in the first millennium BC, not available in the HYDE dataset, suggests that the slowdown in population growth likely started before the first millennium. To find a millennium with slower population growth in this dataset, we have to go back to the fourth millennium BC (5000 BC - 4000 BC).
    All datasets I've looked at report substantial acceleration in population growth with the turn of the second millennium. The Mongol conquests and the Black Death appear to have weighed on population growth from 1200 to 1400, but despite that growth from 1000 to 1500 still averages 0.08%/year, which is faster than the first millennium BC.
    The Lost Decade was a name used in Japan to refer to the ten-year period of economic stagnation following the Japanese stock market crash of 1991. Following this convention, it might be appropriate to call the first millennium "the lost millennium". If we believe in some kind of stochastic hyperbolic growth model of economic history, a la Roodman (2020), this means that humanity got "very unlucky" in the first millennium, possibly due to persistent adverse social shocks.
    This story is purely quantitative, looking at population growth, but it's worth keeping in mind that under Malthusian conditions we expect population growth to track technological progress and capital formation. The more productive we are, the more children people should have until the population increases sufficiently for the marginal product of labor to fall back to subsistence levels. A millennium is more than enough time for Malthusian dynamics to pla...
    5 min
  • LW - Assessment of intelligence agency functionality is difficult yet important by trevor
    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: Assessment of intelligence agency functionality is difficult yet important, published by trevor on August 24, 2023 on LessWrong.
    Summary: When it comes to observing intelligence agencies, it's hard to see the hardened parts and easy to observe the soft corrupt parts. This leads to a bias where very large numbers of people overestimate how prevalent the easily-observed soft and harmless parts are. This can sometimes even result in a dangerous and prevalent estimation, among people whose careers are much further ahead than yours, that the entire intelligence agency is harmless and irrelevant, when it actually isn't. Intelligence agencies are probably a mix of both less-functional, less-relevant parts, and also more-functional, more-relevant parts that have a disproportionately large influence over governments and policies; and it is a mistake to assume that intelligence agencies are homogenously composed of non-functional non-relevant parts that aren't worth paying any attention to, even if such a belief is a popular norm.
    Why intelligence agencies are dangerous
    There are a wide variety of situations where intelligence agencies suddenly becomes relevant, without warning. For example, most or all of the US Natsec establishment might suddenly and unanimously change its stance on Gain of Function research, such as if US-China relations or US-Russian relations once again hit a new 25-year low (which has actually been happening very frequently over the last few years).
    Either the leadership of an agency, or a powerful individual in an agency with authority to execute operations, or a corrupt clique, might personally make a judgement that the best way to expedite or restart GOF research is to target various people who are the most efficient or effective at opposing GOF research.
    This need not be anywhere near the most effective way to expedite or protect GOF research, it just needs to look like that, sufficiently for someone to sign off on that, or even for them to merely thing that it would look good to their boss.
    Competent or technologically advanced capabilities can obviously be mixed with incompetent administration/decisionmaking in the mixed competence model of intelligence agencies. An intelligence agency that is truly harmless, irrelevant, and not worth paying attention to (as opposed to having an incentive to falsely give off the appearance of harmlessness, irrelevance, or not being worth paying attention to) would have to be an intelligence agency that is both technologically unsophisticated and too corrupt for basic functioning, such as running operations.
    This would be an extremely naive belief to have about the intelligence agencies in the US, Russia, and China; particularly the US and China, which have broad prestige, sophisticated technology, and also thriving private sector skill pools to recruit talent from.
    When calculating the expected value from policy advocacy tasks that someone somewhere absolutely must carry out, like pushing sensible policymaking on GOF research that could cause human extinction, many people are currently aware that the risk of that important community disappearing or dissolving substantially reduces the expected value calculations of everything produced by that important community; e.g. a 10% chance of the community ceasing to exist or dissolving reduces the expected value produced by that entire community by something like ~10%.
    Most people I've encountered have in mind a massive totalitarian upheaval, like the ones in the early-mid 20th century, and such an upheaval is a hard boundary between being secure and not being secure. However, in the 21st century, especially after COVID and the 2008 recession, experts and military planners are now more focused on the international balance of power (e.g. the strength of the US, Russia, and ...
    17 min
  • LW - The God of Humanity, and the God of the Robot Utilitarians 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: The God of Humanity, and the God of the Robot Utilitarians, published by Raemon on August 24, 2023 on LessWrong.
    My personal religion involves two gods - the god of humanity (who I sometimes call "Humo") and the god of the robot utilitarians (who I sometimes call "Robutil").
    When I'm facing a moral crisis, I query my shoulder-Humo and my shoulder-Robutil for their thoughts. Sometimes they say the same thing, and there's no real crisis. For example, some naive young EAs try to be utility monks, donate all their money, never take breaks, only do productive things... but Robutil and Humo both agree that quality intellectual world requires slack and psychological health. (Both to handle crises and to notice subtle things, which you might need, even in emergencies)
    If you're an aspiring effective altruist, you should definitely at least be doing all the things that Humo and Robutil agree on. (i.e. get to to the middle point of Tyler Alterman's story here).
    But Humo and Robutil in fact disagree on some things, and disagree on emphasis.
    They disagree on how much effort you should spend to avoid accidentally recruiting people you don't have much use for.
    They disagree on how many high schoolers it's acceptable to accidentally fuck up psychologically, while you experiment with a new program to get them into.
    They disagree on how hard to push yourself to grow better/stronger/wiser/faster, and how much you should sacrifice to do so.
    Humo and Robutil each struggle to understand things differently. Robutil eventually acknowledges you need Slack, but it didn't occur to him initially. His understanding was born in the burnout and tunnel-vision of thousands of young idealists, and Humo eventually (patiently, kindly) saying "I told you so." (Robutil responds "but you didn't provide any arguments about how that maximized utility!". Humo responds "but I said it was obviously unhealthy!" Robutil says "wtf does 'unhealthy' even mean? taboo unhealthy!")
    It took Robutil longer still to consider that perhaps you not only need to prioritize your own wellbeing and your friendships, but that it can be valuable to prioritize them for their own sake, not just as part of a utilitarian calculus, because trying to justify them in utilitarian terms may be a subtly wrong step in the dance that leaves them hollow.
    Humo struggles to acknowledge that if you spend all your time making sure to uphold deontological commitments to avoid harming the people in your care, then this effort is in fact measured in real human beings who suffer and die because you took longer to scale up your program.
    In my headcanon, Humo and Robutil are gods who are old and wise, and they got over their naive struggles long ago. They respect each other as brothers. They understand that each of their perspectives is relevant to the overall project of human flourishing. They don't disagree as much as you'd naively expect, but they speak different languages and emphasize things differently.
    Humo might acknowledge that I can't take care of everyone, or even respond compassionately to all the people who show up in my life who I don't have time to help. But he says so with a warm, mournful compassion, whereas Robutil says in with brief, efficient ruthlessness.
    I find it useful to query them independently, and to imagine the wise version of each of them as best I can - even if my imagining is but a crude shadow of their idealized platonic selves.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org
    4 min
  • EA - New encyclopedia entry on utilitarianism and Christian theology by dominicroser
    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: New encyclopedia entry on utilitarianism and Christian theology, published by dominicroser on August 24, 2023 on The Effective Altruism Forum.
    Theologians have often been very critical about utilitarianism - despite the fact that the first versions of utilitarianism have been developed by Christian thinkers.
    Vesa Hautala and I now got the opportunity to write an encyclopedia article on Christian Theology and Utilitarianism for the St Andrews Encyclopedia of Theology (SAET). SAET is a new project - a theological analogue of the Stanford Encyclopedia of Philosophy.
    This is the first such overview as far as we know. It touches on both historical developments as well as points of overlap and tension. If anyone has criticism to offer of our treatment, we'd happily take it.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org
    1 min
  • EA - Prototype: GPT-powered EA/LW weekly summary by Hamish McDoodles
    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: Prototype: GPT-powered EA/LW weekly summary, published by Hamish McDoodles on August 24, 2023 on The Effective Altruism Forum.
    Zoe Williams used to manually do weekly summaries of the EA Forum and LessWrong, but now she doesn't
    today I strung together a bunch of google apps scripts, google sheets expressions, graphQL queries, and D3.js to automatically extract all the posts on EAF/LW from the last week with >50 karma, summarize them with GPT-4, and assemble the result into HTML with links and stuff
    hard to say what the API usage cost, what with all the tinkering and experimenting, but I reckon it was about $5
    there were a bunch of posts which were too long for the API message length, so as a first whack i just cut stuff out of the middle of the post until it fit (Procrustes style)
    Siao (co-author) was going to help but i finished everything so fast she never got a chance lol
    I haven't spent much time sanity-checking these summaries, but I reckon they're "good enough to be useful"
    They often drop useful details or get the emphasis wrong.
    I haven't seen any outright fabrication.
    Obviously if you have a special interest in some topic these aren't going to substitute for reading the original post.
    the obvious next few steps are:
    automate the actual posting of the summaries.
    does EAF/LW have an API for posting?
    also summarize top comment(s)
    this is kinda hard
    experiment with prompts to see if other summaries are more useful
    also generate a top level summary which gives you like 5 bullet points of the most important things from the forums this week
    feedback that would be useful:
    what would you (personally) like to be different about these summaries? should they be shorter? longer? bullet points? have quotes? fewer entries? more entries?
    leave a comment or DM me or whatever with any old feedback
    oh, and: i originally also got the top posts from the AI alignment forum, but they were all cross posted on lesswrong? is that alwasy true? anyone know?
    EA Forum
    Select examples of adverse selection in longtermist grantmaking
    by Linch
    The author, a volunteer and sometimes contractor for EA Funds' Long-Term Future Fund (LTFF), discusses the pros and cons of diversification in longtermist EA funding. While diversification can increase financial stability, allow for a variety of worldviews, encourage accountability, and provide access to diverse networks, it can also lead to adverse selection, where projects that have been rejected by existing grantmakers are funded by new ones. The author provides examples of such cases and suggests that new grantmakers should be cautious about funding projects that have been rejected by others, but also acknowledges that grantmakers can make mistakes and that a network of independent funders could help ensure that unusual but potentially high-impact projects are not overlooked.
    An Elephant in the Community Building room
    by Kaleem
    The author, a contractor for CEA and employee of EVOps, shares his personal views on the strategies of community building within the Effective Altruism (EA) movement. He identifies two main strategies: Global EA, which aims to spread EA ideas as widely as possible, and Narrow EA, which focuses on influencing a small group of highly influential people. The author argues that community builders and funders should be more explicit about their theory of change for global community building, as there could be significant trade-offs in impact between these two strategies.
    CE alert: 2 new interventions for February-March 2024 Incubation Program
    by CE
    Charity Entrepreneurship has announced two new charity interventions for its February-March 2024 Incubation Program, bringing the total to six. The new interventions include an organization focused on bringing new funding into the animal advocacy movement and an organization providing ...
    21 min
  • LW - Seth Explains Consciousness by Jacob Falkovich
    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: Seth Explains Consciousness, published by Jacob Falkovich on August 24, 2023 on LessWrong.
    The Real Problem
    For as long as there have been philosophers, they loved philosophizing about what life really is. Plato focused on nutrition and reproduction as the core features of living organisms. Aristotle claimed that it was ultimately about resisting perturbations. In the East the focus was less on function and more on essence: the Chinese posited ethereal fractions of qi as the animating force, similar to the Sanskrit prana or the Hebrew neshama. This lively debate kept rolling for 2,500 years - élan vital is a 20th century coinage - accompanied by the sense of an enduring mystery, a fundamental inscrutability about life that will not yield.
    And then, suddenly, this debate dissipated. This wasn't caused by a philosophical breakthrough, by some clever argument or incisive definition that satisfied all sides and deflected all counters. It was the slow accumulation of biological science that broke "Life" down into digestible components, from the biochemistry of living bodies to the thermodynamics of metabolism to genetics. People may still quibble about how to classify a virus that possesses some but not all of life's properties, but these semantic arguments aren't the main concern of biologists. Even among the general public who can't tell a phospholipid from a possum there's no longer a sense that there's some impenetrable mystery regarding how life can arise from mere matter.
    In Being You, Anil Seth is doing the same to the mystery of consciousness. Philosophers of consciousness have committed the same sins as "philosophers of life" before them: they have mistaken their own confusion for a fundamental mystery, and, as with élan vital, they smuggled in foreign substances to cover the gaps. This is René Descartes' res cogitans, a mental substance that is separate from the material.
    This Cartesian dualism in various disguises is at the heart of most "paradoxes" of consciousness. P-zombies are beings materially identical to humans but lacking this special res cogitans sauce, and their conceivability requires accepting substance dualism. The famous "hard problem of consciousness" asks how a "rich inner life" (i.e., res cogitans) can arise from mere "physical processing" and claims that no study of the physical could ever give a satisfying answer.
    Being You by Anil Seth answers these philosophical paradoxes by refusing to engage in all but the minimum required philosophizing. Seth's approach is to study this "rich inner life" directly, as an object of science, instead of musing about its impossibility. After all, phenomenological experience is what's directly available to any of us to observe.
    As with life, consciousness can be broken into multiple components and aspects that can be explained, predicted, and controlled. If we can do all three we can claim a true understanding of each. And after we've achieved it, this understanding of what Seth calls "the real problem of consciousness" directly answers or simply dissolves enduring philosophical conundrums such as:
    What is it like to be a bat?
    How can I have free will in a deterministic universe?
    Why am I me and not Britney Spears?
    Is "the dress" white and gold or blue and black?
    Or at least, these conundrums feel resolved to me. Your experience may vary, which is also one of the key insights about experience that Being You imparts.
    The original photograph of "the dress"
    Seeing a Strawberry
    On a plate in front of you is a strawberry. Inside your skull is a brain, a collection of neurons that have direct access only to the electrochemical state of other neurons, not to strawberries. How does the strawberry out there create the perception of redness in the brain?
    In the common view of perception, red light from the strawberry hi...
    22 min