The Nonlinear Library

The Nonlinear Library

By The Nonlinear FundEducation
Download on the App Store

All The Nonlinear Library episodes

  • EA - Healthier Hens Y1 update including challenges so far and a call for funding by lukasj10
    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: Healthier Hens Y1 update including challenges so far and a call for funding, published by lukasj10 on October 18, 2022 on The Effective Altruism Forum.
    Key points
    Our mission and approach remain largely unchanged
    Our organizational capacity is growing
    We spent most of our Y1 budget on staff, research and travel
    We have made slight, pilot country-informed adjustments to our strategy
    We will ramp up our program work in Y2
    We are finding it rather difficult to fundraise, Y2 budget ($230k) is still not fully covered
    Healthier Hens (HH) received funding from Charity Entrepreneurship, EA Funds and individual donors until now. We need more funding to keep investigating promising dietary interventions to improve the welfare of cage-free hens and engaging cage-free egg farming stakeholders to adopt these interventions. You can donate here.
    Our mission and approach
    Our mission remains to improve the welfare of egg-laying hens kept commercially cage-free. To achieve that, we are investigating promising dietary interventions to reduce the incidence of bone health issues to mitigate hen pain and suffering and engaging cage-free egg farming stakeholders (including farmers, feed mills and regulators) to implement these interventions. Please read our introductory and 6M update posts to learn more about the background of HH.
    Capacity, staff and organizational growth
    We used Y1 to grow in capacity. The corresponding activities involved setting up an advisory board – including African farmed animal welfare (FAW) movement, animal welfare and keel bone damage experts – joining an international FAW association focused on egg-laying hen welfare, visiting egg farms and feed producers in Kenya, and running initial animal welfare workshops. This has permitted us to be well aware of ongoing cage-free campaigns and existing hen welfare interventions, learn what practical obstacles the industry faces, and what knowledge/capacity gaps must be addressed to facilitate our program work in improving hen welfare.
    After choosing our pilot country of operations, our main task was to hire a full-time Country Manager to facilitate our efforts on the ground. Our team has also seen two Research Interns joining our efforts to build knowledge capacity on how hen nutrition relates to welfare issues we are seeing on farms. Looking forward, we will recruit Data Collection and Operation Interns in Kenya, remote Research Interns and a hen welfare/nutrition specialist during Y2, which should bring the total of paid HH staff up to 8.
    We had a total operational budget of $159k for Y1. Its break-down can be seen below. The majority of our funds were used for Co-founder and staff salaries, research (feed trial in Switzerland) and travel expenses (mainly related to, initially, country scoping and on-the-ground Kenya activities). Y2 will see more of our budget accounting for non Co-founder staff expenses and program expenses. The total estimated Y2 budget is currently at $230k, with $50k raised so far.
    Strategic updates
    Our country selection process took a significant part of Y1 to carry out (read more about it here). We are building up knowledge and networks within the local cage-free egg production and feed manufacturing industries via co-founder country visits and the work led by our Country Manager. The need to build reputation and establish actively connected networks within the different communities became apparent as a necessity to initiate pilot work. This is why we are ramping up the farmer training portion of our work while the other research activities are ongoing. The workshops enable us to achieve the following goals in our target counties:
    Gauge farmer welfare awareness and identify gaps crucial for ensuring hen bone development, health and integrity.
    Build relationships with the county level Veterinary Dir...
    10 min
  • EA - Effective Altruism's Implicit Epistemology by Violet Hour
    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: Effective Altruism's Implicit Epistemology, published by Violet Hour on October 18, 2022 on The Effective Altruism Forum.
    Cross-posted from The Violet Hour.
    Note: This is a piece on the sociology of (longtermist) EA, aimed to be accessible to a wider audience. While I intend the post to also interest more immersed EAs, many readers here should feel comfortable heavily skimming the introduction, and likewise skimming (or skipping) Section 2, which provides an informal exposition of Subjective Bayesianism and Expected Value Theory.
    The appendix (~1.1k words) offers a preliminary discussion of some philosophical issues related to this post.
    1. Introduction
    The future might be very big, and we might be able to do a lot, right now, to shape it.
    You might have heard of a community of people who take this idea pretty seriously — effective altruists, or ‘EAs’. If you first heard of EA a few years ago, and haven’t really followed it since, then you might be pretty surprised at where we’ve ended up.
    Currently, EA consists of a variety of professional organizations, researchers, and grantmakers, all with (sometimes subtly) different approaches to doing the most good possible. Organizations which, collectively, donate billions of dollars towards interventions aiming to improve the welfare of conscious beings. In recent years, the EA community has shifted its priorities towards an idea called longtermism — very roughly, the idea that we should primarily focus our altruistic efforts towards shaping the very long-run future. Like, very long-run. At least thousands of years. Maybe more.
    (From hereon, I’ll use ‘EA’ to talk primarily about longtermist EA. Hopefully this won’t annoy too many people).
    Anyway, longtermist ideas have pushed EA to focus on a few key cause areas — in particular, ensuring that the development of advanced AI is safe, preventing the development of deliberately engineered pandemics, and (of course) promoting effective altruism itself. I've been part of this community for a while, and I've often found outsiders bemused by some of our main priorities. I've also found myself puzzled by this bemusement. The explicit commitments undergirding (longtermist) EA, as many philosophers involved with EA remind us, are really not all that controversial. And those philosophers are right, I think. Will MacAskill, for instance, has listed the following three claims as forming the basic, core argument behind longtermism.
    (1) Future people matter morally.
    (2) There could be enormous numbers of future people.
    (3) We can make a difference to the world they inhabit.
    Together, these three claims all seem pretty reasonable. With that in mind, we’re left with a puzzle: given that EA’s explicitly stated core commitments are not that weird, why, to many people, do EA’s explicit, practical priorities appear so weird?
    In short, my answer to this puzzle claims that EA’s priorities emerge, to a large extent, from EA’s unusual epistemic culture. So, in this essay, I’ll attempt to highlight the sociologically distinctive norms EAs adopt, in practice, concerning how to reason, and how to prioritize under uncertainty. I’ll then claim that various informal norms, beyond some of the more explicit philosophical theories which inspire those norms, play a key role in driving EA’s prioritization decisions.
    2. Numbers, Numbers, Numbers
    Suppose I asked you, right now, to give me a probability that Earth will experience an alien invasion within the next two weeks.
    You might just ignore me. But suppose you‘ve come across me at a party; the friend you arrived with is conspicuously absent. You look around, and, well, the other conversations aren’t any better. Also, you notice that the guy you fancy is in the corner; every so often, you catch him shyly glancing at us. Fine, you think, I could be crazy, but ...
    49 min
  • AF - An Extremely Opinionated Annotated List of My Favourite Mechanistic Interpretability Papers by Neel Nanda
    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: An Extremely Opinionated Annotated List of My Favourite Mechanistic Interpretability Papers, published by Neel Nanda on October 18, 2022 on The AI Alignment Forum.
    Introduction
    This is an extremely opinionated list of my favourite mechanistic interpretability papers, annotated with my key takeaways and what I like about each paper, which bits to deeply engage with vs skim (and what to focus on when skimming) vs which bits I don’t care about and recommend skipping, along with fun digressions and various hot takes.
    This is aimed at people trying to get into the field of mechanistic interpretability (especially Large Language Model (LLM) interpretability). I’m writing it because I’ve benefited a lot by hearing the unfiltered and honest opinions from other researchers, especially when first learning about something, and I think it’s valuable to make this kind of thing public! On the flipside though, this post is explicitly about my personal opinions - I think some of these takes are controversial and other people in the field would disagree.
    The four top level sections are priority ordered, but papers within each section are ordered arbitrarily - follow your curiosity
    Priority 1: What is Mechanistic Interpretability?
    Circuits: Zoom
    In
    Sets out the circuits research agenda, and is a whirlwind overview of progress in image circuits
    This is reasonably short and conceptual (rather than technical) and in my opinion very important, so I recommend deeply engaging with all of it, rather than skimming.
    The core thing to take away from it is the perspective of networks having legible(-ish) internal representations of features, and that these may be connected up into interpretable circuits. The key is that this is a mindset for thinking about networks in general, and all the discussion of image circuits is just grounding in concrete examples.
    On a deeper level, understanding why these are important and non-trivial claims about neural networks, and their implications.
    In my opinion, the circuits agenda is pretty deeply at the core of what mechanistic interpretability is. It’s built on the assumption that there is some legible, interpretable structure inside neural networks, if we can just figure out how to reverse engineer it. And the core goal of the field is to find what circuits we can, build better tools for doing so, and do the fundamental science of figuring out which of the claims about circuits are actually true, which ones break, and whether we can fix them.
    An important note is that mechanistic interpretability is an extremely young field and this was written 2.5 years ago - I take the specific claims in this article as a starting point, not as the definitive grounding of what the field must believe.
    Meta: The goal of reading this is to understand what the fundamental mindset and worldview being defended here is. The goal is not necessarily to leave feeling convinced that these claims are true, or that the article adequately justifies them. That’s what the rest of the papers in here are for!
    A useful thing to reflect on is what the world would look like if the claims were and were not true - what evidence could you see that might convince you either way? These are definitely not obviously true claims!
    A Mathematical Framework for Transformer
    Circuits
    The point of this is to explain how to conceptually break down a transformer into individually understandable pieces.
    Deeply engage with:
    All the ideas in the overview section, especially:
    Understanding the residual stream and why it’s fundamental.
    The notion of interpreting paths between interpretable bits (eg input tokens and output logits) where the path is a composition of matrices and how this is different from interpreting every intermediate activations
    And understanding attention heads: what a QK and OV matrix is, how a...
    20 min
  • AF - Distilled Representations Research Agenda by Hoagy
    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: Distilled Representations Research Agenda, published by Hoagy on October 18, 2022 on The AI Alignment Forum.
    Introduction
    I’ve recently been given funding from the Long Term Future Fund to develop work on an agenda I'll tentatively call Distilled Representations, and I'll be working on this full-time over the next 6 months with Misha Wagner (part time).
    We're working on a way of training autoencoders so that they can only represent information in certain ways - ways that we can define in a flexible manner.
    It works by training multiple autoencoders to encode a set of objects, while for some objects defining a preferred representation that the autoencoders are encouraged to encode the objects as. We then distill these multiple autoencoders into single autoencoder which encodes only that information which is encoded in the same way across the different autoencoders. If we are correct, this new autoencoder should only encode information using the preferred strategy. Vitally, this can be not just the original information in the preferred representations, but also information represented by generalizations of that encoding strategy.
    It is similar to work such as Concept Bottleneck Models but we hope the distillation from multiple models should allow interpretable spaces in a much broader range of cases.
    The rest of this post gives more detail of the intuition that we hope to build into a useful tool, some toy experiments we’ve performed to validate the basic concepts, the experiments that we hope to build in the future, and the reasons we hope it can be a useful tool for alignment.
    We'd like to make sure we understand what similar work has been done and where this work could be useful. If you're familiar with disentangled representations, or interpretability tools more generally, we're interested in having a chat. You can reach me here on LessWrong or at [email protected].
    Previous versions of similar ideas can be found in my ELK submission and especially Note-taking Without Hidden Messages.
    Intuition
    The intuition that this work builds on is the following:
    With neural networks, the meanings of the weights and activations are usually opaque but we're often confident about the kind of thing that the network must be representing, at least for some cases or parts of the input distribution.
    In those cases where we understand what the network is representing, we can condense this understanding into a vector, thus defining a 'preferred representation' which encapsulates that knowledge.
    We can compress the NN's state with an autoencoder, while in those cases with preferred representations, encouraging the encoding to be as close as possible to the preferred representation.
    We expect that this running this compression results in the known information being compressed in the manner specified by the preferred representations, while other important information also being snuck in wherever possible.
    If we then train multiple encoder/decoder systems, they will use the preferred representation, but also will use generalizations of the preferred representations.
    Additional info that is not a generalization of the preferred representation scheme will also be encoded, but the encoding scheme for additional information will vary between different encoder/decoder pairs.
    Using methods such as retraining a new encoder to encode for randomly shuffled decoders at each batch, we can create an encoder that uses a generalization of our preferred encoding scheme, without containing additional, misleading information.
    There are quite a few leaps in this reasoning, and we view the key assumptions / hypotheses to be tested as the following:
    In relevant situations we can define preferred representations.
    We can force encoders to use not just these representations but meaningful generalizations ...
    14 min
  • EA - Careers in medicine - a new path profile from Probably Good and High Impact Medicine by Probably Good
    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: Careers in medicine - a new path profile from Probably Good and High Impact Medicine, published by Probably Good on October 18, 2022 on The Effective Altruism Forum.
    The teams at Probably Good and High Impact Medicine (Hi-Med) are excited to co-publish a new path profile on careers in medicine! You can find the profile over at both the Probably Good and Hi-Med websites.
    Summary
    The path profile is an introductory guide to impactful careers within the medical space, primarily aimed at existing doctors and medical students who may not yet be familiar with EA principles.
    One goal for this profile is to provide actionable advice for people who may be more resistant to wholesale career changes. We understand that large career changes can be daunting, especially for people who have spent many years aspiring (and training) to become a doctor. With this in mind, we want to give information for people who may be strongly inclined towards a career in clinical medicine, and want to find ways to increase their impact. The article suggests ways in which medics might increase their impact within clinical careers – for example, by pursuing high-paying specialities and donating to effective charities – as well as pointing out where doctors may be able to get involved with high-impact activities alongside their clinical duties.
    However, the profile also makes clear that the highest impact choices for many people will likely lie outside clinical work, and we therefore highlight a few of the most salient high-impact alternatives both in the medical field more broadly, as well as outside medicine, for someone with a medical background. The career paths we discuss include medical research, public health, biosecurity, and nonprofit entrepreneurship.
    Below, we’ve included the full introduction to the profile as well as a couple of representative excerpts.
    Introduction
    The medical profession attracts huge numbers of altruistically-minded people who want to make a difference with their careers. One recent survey, which asked for the career aspirations of over 600,000 teenagers across the world, found that over 10% of them wanted to become doctors. This is great – doctors are vital, and the world is much better because of them.
    But if you’ve gone through medical school, you’ll be familiar with the concept of triage – that is, treating patients according to the urgency of their needs and your ability to help. Sometimes this means making tough calls about who to help first and how to help them. Though triage can be emotionally challenging, it’s vital for saving lives.
    We think it’s possible to apply a kind of triage to the world’s problems – working out which are most urgent and deciding on the most effective ways to solve them. In doing so, we may be able to identify career choices that allow us to achieve even more good than we might do otherwise.
    This kind of career-focused triage can be applied here in two ways. First, we can look for the highest priority paths within clinical medicine, by searching for where doctors are most needed, and what they can do to help most. Secondly, we can look at alternatives to conventional clinical careers to see if we can have more impact on the world through another career route. Perhaps surprisingly, it turns out that there are a number of career paths in which you might be able to do more good than you would as a doctor, either because these paths focus on more pressing problems, or provide larger-scale solutions than treating patients individually.
    In this profile, we'll highlight a few of these possible high-impact career routes – including biosecurity, nonprofit entrepreneurship, medical research, and public health. If one or more of these options could be a good fit for you, then you could have multiple times the impact in these careers than you might...
    17 min
  • LW - Is GitHub Copilot in legal trouble? by tcelferact
    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: Is GitHub Copilot in legal trouble?, published by tcelferact on October 18, 2022 on LessWrong.
    This is basically a crosspost for/. I noticed that some folks in California are considering a lawsuit against Microsoft/OpenAI.
    tl;dr:
    Copilot is trained on open source software.
    Copilot doesn't respect the licensing agreements of that software.
    Copilot doesn't have a clear fair use argument for doing so.
    By accepting copilot suggestions, you are potentially violating licensing agreements yourself.
    Sections of particular interest:
    [W]e inquired pri­vately with Fried­man and other Microsoft and GitHub rep­re­sen­ta­tives in June 2021, ask­ing for solid legal ref­er­ences for GitHub’s pub­lic legal posi­tions . They pro­vided none.
    Software Freedom Conservancy
    “You are respon­si­ble for ensur­ing the secu­rity and qual­ity of your code. We rec­om­mend you take the same pre­cau­tions when using code gen­er­ated by GitHub Copi­lot that you would when using any code you didn’t write your­self. These pre­cau­tions include rig­or­ous test­ing, IP [(= intel­lec­tual prop­erty)] scan­ning [my emphasis], and track­ing for secu­rity vul­ner­a­bil­i­ties.”
    #using-github-copilot
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    2 min
  • AF - A conversation about Katja's counterarguments to AI risk by Matthew Barnett
    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 conversation about Katja's counterarguments to AI risk, published by Matthew Barnett on October 18, 2022 on The AI Alignment Forum.
    This post is a transcript of a conversation between Ege Erdil and Ronny Fernandez, recorded by me. The participants talked about a recent post by Katja Grace that presented many counterarguments to the basic case for AI x-risk. You might want to read that post first.
    As it was transcribed automatically by Whisper, along with some light editing on top, there might be many mistakes in the transcript. You can also find the audio file here, but note that the first five lines of the dialogue do not appear in the audio.
    Ronny Fernandez wants to make it clear that he did not know the conversation would eventually be published. (But all participants have consented to it being published).
    Ege ErdilYou know about Epoch, right?
    Ronny FernandezNot really, actually.
    Ege ErdilOK, so Epoch is a recently established organization that focuses on forecasting future developments in AI. And Matthew is currently working there and hosted this post, like, with Slack. So there's some discussion of that.
    Ronny FernandezGotcha.
    Ege ErdilAnd he basically said that he thought that it was a good post. We didn't. I mean, it's not just me. There's also someone else who thought that it wasn't really very good. And that's just something he said, oh, you know, I should talk to Ronny because he also thinks that the post was good.
    Ronny FernandezYeah, sweet. So question I want to ask is, do you think the post wasn't good because you think that the summary of the basic argument for existential risk from superintelligence is not the right argument to be criticizing? Or do you think the post wasn't good because the criticisms weren't good of that argument or for some other reason?
    Ege ErdilI think it's both. I think there's somewhat of criticisms are, I think, quite poor. But I also think the summary of the argument is not necessarily that great. So yeah. So we could also go through the post and order that arguments are given. Or I could just talk about the parts that we talked about.
    Ronny FernandezI'll pull up the post now. No, I think it'd be great to take a look at the argument and talk about how you would repartition the premises or change the premises or whatever.
    Ege ErdilYeah. So let me first go to the last section, because I think it is actually interesting in the sense that the given arguments could in principle apply to corporations. And the post points this out. This is an example. It's been given before. And it says this argument proves too much, because it also proves that we have to be very worried about corporations. There's a certain logic to that. But I think what is overlooked in the post is that the reason we don't worry much about corporations is that corporations are very bad at coordinating, even internally, let alone across different corporations cooperating. So I think the main reason we should be worried about AI, and there are some scenarios in which takeoff is very fast, and one guy builds something in his basement, and that takes over the world. But I think those scenarios are pretty unlikely.
    But I do think if you have that kind of view, the post is going to be pretty unconvincing to you, because it doesn't really argue against that situation. But I don't really believe that's plausible.
    For me, the big problem is that the post does not focus on the fact that it will be much, much easier for AIs that have different goals to coordinate than it should be for humans. Because AIs can have very straightforward protocols that, especially if they are goal-directed in the sense defined by the post, you just take two goals and merge them according to some weights, and you have this new AI that's bigger. And it's a mixture of the two things that was befo...
    49 min
  • EA - EA Funds has a Public Grants Database by calebp
    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 Funds has a Public Grants Database, published by calebp on October 18, 2022 on The Effective Altruism Forum.
    We have just launched a public grants database which contains information on all of our public grants .
    We currently include details of the fund that made the grant, a project summary, the grantee and the payout amount.
    We think it is valuable to show the EA community and our donors what kinds of projects we end up funding to increase accountability and encourage more excellent projects. Prospective grantees often ask me what types of projects we fund, and I hope this will make it easier for them to come up with ideas for projects and know whether their project ideas are in scope.
    We previously wrote public payout reports where our fund managers would outline their reasoning for making a grant. I think that the public payout reports were excellent. Still, they were time-consuming for our fund managers, and I would prefer them to spend their limited capacity evaluating grants . We were also much slower than I would have liked with releasing them, and I hope the database will stay much more up-to-date with our grantmaking (with entries generally added soon after we make payouts).
    Please note that it is currently still being tested, so some grant information may be incorrect.
    One idea is to ask fund managers to highlight a few grants each month and explain why they were excited about them to encourage more excellent projects and give insight into our grantmaking. They would not be comprehensive like our old reports, but I think our public grants database and these highlights reports (when we have time to write them) will get most of the value of the old reports while saving us substantial time.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    2 min
  • AF - Science of Deep Learning - a technical agenda by Marius Hobbhahn
    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: Science of Deep Learning - a technical agenda, published by Marius Hobbhahn on October 18, 2022 on The AI Alignment Forum.
    I have written down a long list of alignment ideas that I’d be interested in working on. The ideas roughly boil down to “To make progress on alignment, we need to understand Deep Learning models and the process by which they arrive at their final parameters in much more detail than we currently do”.
    I’m not the first person to think of most of these ideas and it builds on a lot of other people’s work. You should think of it more as a collection of existing resources and ideas than a new agenda. I also haven’t come up with the term “Science of Deep Learning”. I have already heard it being used by multiple people within the alignment community and many researchers are already working on parts of this agenda.
    I obviously don’t own this agenda. The questions in this list are sufficient to keep hundreds of researchers busy for a while, so feel free to hop on. If you’re interested in collaborating just reach out.
    Some of this research has the potential to increase capabilities more than alignment and the results should, in some cases, be kept private and only discussed with a small group of trusted peers. However, I think that most of the projects have a “defender’s advantage”, i.e. they increase alignment more than capabilities.
    Whenever possible, Science of DL projects should have a direct benefit for alignment but I think our current understanding of DL is so bad that just increasing our general understanding seems like a good start.
    Here is the link to the full version (comments are on, please don’t abuse it):
    The rest of this post is an overview copied from the doc. Feedback is welcome.
    Overview - Science of Deep Learning
    By Science of DL, I roughly mean “understanding DL systems and how they learn concepts” better. The main goal is to propose a precise and testable hypothesis related to a phenomenon in DL and then test and refine it until we are highly confident in its truth or falsehood. This hypothesis could be about how NNs behave on the neuron level, the circuit level, during training, during fine-tuning, etc. This research will almost surely at some point include mechanistic interpretability but it is not limited to it. The refined statement after investigation can but doesn’t have to be of mathematical form as long as it is unambiguous and can be tested, i.e. two people could agree on an experiment that would provide evidence for or against the statement and then run it.
    How this could look in practice
    The details would obviously differ from project to project but on a high level I imagine it to look roughly like this
    Pick an interesting concept found in deep learning, e.g. grokking, the lottery ticket hypothesis, adversarial examples or the emergence of 2-digit addition in LLMs. Optimally, the concept is safety-related but especially in the beginning, just increasing general understanding seems more important than the exact choice of topic.
    Try to understand high-level features of the phenomenon, e.g. under which conditions this concept arises, which NNs show it and which ones don’t, in which parts of the networks it arises, when during training it arises, etc. This likely includes retraining the network under different conditions with different hyperparameters, number of parameters, etc. and monitoring meaningful high-level statistics related to the concept, e.g. monitor the validation loss to see when the model starts to grok.
    Zoom in: try to understand what happens on a low level, e.g. use mechanistic interpretability tools to investigate the neurons/activations or use other techniques to form a hypothesis of how this specific part of the network works. In the optimal case, we would be able to describe the behavior very precisely, e.g. ...
    8 min
  • EA - Decision theory does not imply that we get to have nice things by So8res
    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: Decision theory does not imply that we get to have nice things, published by So8res on October 18, 2022 on The Effective Altruism Forum.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    1 min