The Nonlinear Library

The Nonlinear Library

By The Nonlinear FundEducation
Download on the App Store

All The Nonlinear Library episodes

  • LW - Take 4: One problem with natural abstractions is there's too many of them. by Charlie Steiner
    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: Take 4: One problem with natural abstractions is there's too many of them., published by Charlie Steiner on December 5, 2022 on LessWrong.
    As a writing exercise, I'm writing an AI Alignment Hot Take Advent Calendar - one new hot take, written every day for 25 days. Or until I run out of hot takes.
    Everyone knows what the deal with natural abstractions is, right? Abstractions are regularities about the world that are really useful for representing its coarse grained behavior - they're building blocks for communicating, compressing, or predicting information about the world. An abstraction is "natural" if it's so easy to learn, and so broadly useful, that most right-thinking agents will have it as part of their toolbox of abstractions.
    The dream is to use natural abstractions to pick out what we want from an AI. Suppose "human values" are a natural abstraction: then both humans and a world-modeling AI would have nearly the exact same human values abstraction in their toolboxes of abstractions. If we can just activate the AI's human values abstraction, we can more or less avoid misalignment between what-humans-are-trying-to-pick-out and what-abstraction-the-AI-takes-as-its-target.
    One might think that the main challenge to this plan would be if there are too few natural abstractions. If human values (or agency, or corrigibility, or whatever nice thing you want to target) aren't a natural abstraction, you lose that confidence that the human and the AI are pointing at the same thing. But it's also a challenge if there are too many natural abstractions.
    Turns out, humans don't just have one abstraction that is "human values," they have a whole lot of 'em. Humans have many different languages / ontologies we use to talk about people, and these use different abstractions as building blocks. More than one of these abstractions gets called "human values," but they're living in different ontologies / get applied in different contexts.
    If none of these abstractions we use to talk about human values are natural, then we're back to the first problem. But if any of them are natural, it seems just as plausible that nearly all of them are. Abstractions don't even have to be discrete - it's perfectly possible to have a continuum.
    This complicates the easy alignment plan, because it means that the structure of the world is merely doing most of the work for us rather than almost all of the work. The bigger the space of semantically-similar natural abstractions you have to navigate, the more you have to be careful about your extensional definitions, and the higher standards you have to have for telling good from bad results.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    3 min
  • LW - Who are some prominent reasonable people who are confident that AI won't kill everyone? by Optimization Process
    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: Who are some prominent reasonable people who are confident that AI won't kill everyone?, published by Optimization Process on December 5, 2022 on LessWrong.
    Bounty: $30 for each link that leads to me reading/hearing ~500 words from a Respectable Person arguing, roughly, "accelerating AI capabilities isn't bad," and me subsequently thinking "yeah, that seemed pretty reasonable." For example, linking me to nostalgebraist or OpenAI's alignment agenda or this debate. Total bounty capped at $600, first come first served. All bounties (incl. the total-bounty cap) doubled if, by Jan 1, I can consistently read people expressing unconcern about AI and not notice a status-yuck reaction.
    Context: I notice that I've internalized a message like "thinking that AI has a <1% chance of killing everyone is stupid and low-status." Because I am a monkey, this damages my ability to consider the possibility that AI has a <1% chance of killing everyone, which is a bummer, because my beliefs on that topic affect things like whether I continue to work at my job accelerating AI capabilities.
    I would like to be able to consider that possibility rationally, and that requires neutralizing my status-yuck reaction. One promising-seeming approach is to spend a lot of time looking at lots of of high-status monkeys who believe it!
    Bounty excludes things I've already seen, and things I would have found myself based on previous recommendations for which I paid bounties (for example, other posts by the same author on the same web site).
    Lest ye worry that [providing links to good arguments] will lead to [me happily burying my head in the sand and continuing to hasten the apocalypse] -- a lack of links to good arguments would move much more of my probability-mass to "Less Wrong is an echo chamber" than to "there are basically no reasonable people who think advancing AI capabilities is good."
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    3 min
  • EA - Should strong longtermists really want to minimize existential risk? by tobycrisford
    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: Should strong longtermists really want to minimize existential risk?, published by tobycrisford on December 4, 2022 on The Effective Altruism Forum.
    Strong longtermists believe there is a non-negligible chance that the future will be enormous. For example, earth-originating life may one day fill the galaxy with 1040 digital minds. The future therefore has enormous expected value, and concern for the long-term should almost always dominate near-term considerations, at least for those decisions where our goal is to maximize expected value.
    It is often stated that strong longtermism reduces in practice to the goal: “minimize existential risk at all costs”. I argue here that this is inaccurate. I claim that a more accurate way of summarising the strong longtermist goal is: “minimize existential risk at all costs conditional on the future possibly being very big”. I believe the distinction between these two goals has important practical implications. The strong longtermist goal may actually conflict with the goal of minimizing existential risk unconditionally.
    In the next section I describe a thought experiment to demonstrate my claim. In the following section I argue that this is likely to be relevant to the actual world we find ourselves in. In the final section I give some concluding remarks on what we should take away from all this.
    The Anti-Apocalypse Machine
    The Earth is about to be destroyed by a cosmic disaster. This disaster would end all life, and snuff out all of our enormous future potential.
    Fortunately, physicists have almost settled on a grand unified theory of everything that they believe will help them build a machine to save us. They are 99% certain that the world is described by Theory A, which tells us we can be saved if we build Machine A. But there is a 1% chance that the correct theory is actually Theory B, in which case we need to build Machine B. We only have the time and resources to build one machine.
    It appears that our best bet is to build Machine A, but there is a catch. If Theory B is true, then the expected value of our future is many orders of magnitude larger (although it is enormous under both theories). This is because Theory B leaves open the possibility that we may one day develop slightly-faster-than-light travel, while Theory A being true would make that impossible.
    Due to the spread of strong longtermism, Earth's inhabitants decide that they should build Machine B, acting as if the speculative Theory B is correct, since this is what maximizes expected value. Extinction would be far worse in the Theory B world than the Theory A world, so they decide to take the action which would prevent extinction in that world. They deliberately choose a 99% chance of extinction over a 1% chance, risking all of humanity, and all of humanity's future potential.
    The lesson here is that strong longtermism gives us the goal to minimize existential risk conditional on the future possibly being very big, and that may conflict with the goal to minimize existential risk unconditionally.
    Relevance for the actual world
    The implication of the above thought experiment is that strong longtermism tells us to look at the set of possible theories about the world, pick the one in which the future is largest, and, if it is large enough, act as if that theory were true. This is likely to have absurd consequences if carried to its logical conclusion, even in real world cases. I explore some examples in this section.
    The picture becomes more confusing when you consider theories which permit the future to have infinite value. In Nick Beckstead's original thesis, On the Overwhelming Importance of Shaping the Far Future, he explicitly singles out infinite value cases as examples of where we should abandon expected value maximization, and switch to using a more timid deci...
    7 min
  • EA - Do Brains Contain Many Conscious Subsystems? If So, Should We Act Differently? by Bob Fischer
    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: Do Brains Contain Many Conscious Subsystems? If So, Should We Act Differently?, published by Bob Fischer on December 5, 2022 on The Effective Altruism Forum.
    Key Takeaways
    The Conscious Subsystems Hypothesis (“Conscious Subsystems,” for short) says that brains have subsystems that realize phenomenally conscious states that aren’t accessible to the subjects we typically associate with those brains—namely, the ones who report their experiences to us.
    Given that humans’ brains are likely to support more such subsystems than animals’ brains, EAs who have explored Conscious Subsystems have suggested that it provides a reason for risk-neutral expected utility maximizers to assign more weight to humans relative to animals.
    However, even if Conscious Subsystems is true, it probably doesn’t imply that risk-neutral expected utility maximizers ought to allocate neartermist dollars to humans instead of animals. There are three reasons for this:
    If humans have conscious subsystems, then animals probably have them too, so taking them seriously doesn’t increase the expected value of, say, humans over chickens as much as we might initially suppose.
    Risk-neutral expected utility maximizers are committed to assumptions—including the assumption that all welfare counts equally, whoever’s welfare it is—that support the conclusion that the best animal-focused neartermist interventions (e.g., cage-free campaigns) are many times better than the best human-focused neartermist interventions (e.g., bednets).
    Independently, note that the higher our credences in the theories of consciousness that are most friendly to Conscious Subsystems, the higher our credences ought to be in the hypothesis that many small invertebrates are sentient. So, insofar as we’re risk-neutral expected utility maximizers with relatively high credences in Conscious Subsystems-friendly theories of consciousness, it’s likely that we should be putting far more resources into investigating the welfare of the world’s small invertebrates.
    We assign very low credences to claims that ostensibly support Conscious Subsystems.
    The appeal of the idea that standard theories of consciousness support Conscious Subsystems may be based on not distinguishing (a) theories that are just designed to make predictions about when people will self-report having conscious experiences of a certain type (which may all be wrong, but have whatever direct empirical support they have) and (b) theories that are attempts to answer the so-called “hard problem” of consciousness (which only have indirect empirical support and are far more controversial).
    Standard versions of functionalism say that states are conscious when they have the right relationships to sensory stimulations, other mental states, and behavior. But it’s highly unlikely that many groups of neurons stand in the correct relationships, even if they perform functions that, in the abstract, seem as complex and sophisticated as those performed by whole brains.
    Ultimately, we do not recommend acting on Conscious Subsystems at this time.
    Introduction
    This is the fifth post in the Moral Weight Project Sequence. The aim of the sequence is to provide an overview of the research that Rethink Priorities conducted between May 2021 and October 2022 on interspecific cause prioritization—i.e., making resource allocation decisions across species. The aim of this post is to assess a hypothesis that's been advanced by several members of the EA community: namely, that brains have subsystems that realize phenomenally conscious states that aren’t accessible to the subjects we typically associate with those brains (i.e., the ones who report their experiences to us; see, e.g., Tomasik, 2013-2019, Shiller, 2016, Muehlhauser, 2017, Shulman, 2020, Crummett, 2022).
    If there are such states, then we might think that ...
    43 min
  • EA - What specific changes should we as a community make to the effective altruism community? [Stage 1] by Nathan Young
    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 specific changes should we as a community make to the effective altruism community? [Stage 1], published by Nathan Young on December 5, 2022 on The Effective Altruism Forum.
    I'm want to run the listening exercise I'd like to see.
    Get popular suggestions
    Run a polis poll
    Make a google doc where we research consensus suggestions/ near consensus/consensus for specific groups
    Poll again
    Stage 1
    Give concrete suggestions for community changes. 1 - 2 sentences only.
    Upvote if you think they are worth putting in the polis poll and agreevote if you think the comment is true.
    Agreevote if you think they are well-framed.
    Aim for them to be upvoted. Please add suggestions you'd like to see.
    I'll take the top 20 - 30
    I will delete/move to comments top-level answers that are longer than 2 sentences.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    2 min
  • LW - Nook Nature by Duncan Sabien
    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: Nook Nature, published by Duncan Sabien on December 5, 2022 on LessWrong.
    Style of post: Alexanderian meander (though with (at best) a tenth of the artistry and thoroughness/rigor). There's a Thing here and it has been tickling at my brain for ages and I have despaired of being able to clarify it all on my own, so I'm giving you my unclarified thoughts instead. More of an extended koan than a proper essay.
    An acquaintance at EA Global suggested that I investigate anagram potential before settling on a possible name for my future child.
    This was a cool idea, so I went online with my frontrunner to see what the possibilities were.
    There were over 111,000 of them.
    If I had made some modest change, say replacing "Elizabeth" with "Frederick" while leaving the other names alone, there would have been a comparable number of completely different ones.
    There's something blank about infinity, or really large numbers like "one billion."
    But the explosion of possibility somehow felt more viscerally real when it was like expanding one Workflowy node to see a hundred thousand possibilities, and expanding the next Workflowy node to see a hundred thousand different possibilities, and knowing that the next node would contain a hundred thousand new possibilities still.
    Most people are familiar with the Youtube videos showing an infinite zoom into the Mandelbrot set, and how you keep finding complexity as you go deeper and deeper.
    But it's interesting (and often underemphasized, in my opinion) that those videos show you the beauty and complexity of one single path, and if you were at any point to go left instead of right, or shift the trajectory of your zooming by some tiny amount in the hundredth decimal place, you would get another, extremely different path of equivalent complexity and beauty.
    This is one of the talking points that has come up in my chocolate tastings. I sometimes like to mention that yeah, of course, there's fantastic, staggering, unmanageable complexity everywhere around us all the time, and of course we have to mute and muffle and abstract most of it away or we would be fully paralyzed and unable to comprehend anything or take any action whatsoever
    (Something something autism, something something processing disorders)
    but it's nice, when someone highlights a particular nook in the infinite fractal chaos as being pleasant and interesting and worth some extra lingering, to stop glossing over all the detail, or at least to gloss over it a little bit less. When I lead people through a chocolate tasting, most of what is special isn't the chocolate itself; it's that they're simply actually processing a bunch of sense data that, by default, would have been ignored, because it has to be ignored, because you have to ignore 99.999% of what you perceive to be functional at all.
    An old puzzle used to be: if the universe is effectively infinite in extent, and per-unit-of-arc every star in the sky is approximately as bright as the sun, and there are stars in every direction ... why isn't the sky a uniform shining white?
    There are several things wrong with the "if," but nevertheless it was useful for astronomers and physicists to notice that, if stars cluster, such that a galaxy might be hidden behind a single nearby star, and might itself obscure a supercluster yet further away, then we could indeed have infinite stars in every direction and still have most of the sky be black. A sort of inverted Cantor dust effect.
    Speaking of fractals, I keep getting this image of something like an infinite anthill, and I don't have anything to draw with so I'll try words and maybe it'll work.
    But like. Okay, so when I was born I (basically) went from a womb to a bedroom. And as I grew, the bubble of my awareness expanded to a house, and then to a whole property, and then to a neighborhood, a...
    16 min
  • EA - Race to the Top: Benchmarks for AI Safety by isaduan
    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: Race to the Top: Benchmarks for AI Safety, published by isaduan on December 4, 2022 on The Effective Altruism Forum.
    This is an executive summary of a blog post. Read the full texts here.
    Summary
    Benchmarks support the empirical, quantitative evaluation of progress in AI research. Although benchmarks are ubiquitous in most subfields of machine learning, they are still rare in the subfield of AI safety.
    I argue that creating benchmarks should be a high priority for AI safety. While this idea is not new, I think it may still be underrated. Among other benefits, benchmarks would make it much easier to:
    track the field’s progress and focus resources on the most productive lines of work;
    create professional incentives for researchers - especially Chinese researchers - to work on problems that are relevant to AGI safety;
    develop auditing regimes and regulations for advanced AI systems.
    Unfortunately, we cannot assume that good benchmarks will be developed quickly enough “by default." I discuss several reasons to expect them to be undersupplied. I also outline actions that different groups can take today to accelerate their development.For example, AI safety researchers can help by:
    directly trying their hand at creating safety-relevant benchmarks;
    clarifying certain safety-relevant traits (such as “honesty” and “power-seekingness”) that it could be important to measure in the future;
    building up relevant expertise and skills, for instance by working on other benchmarking projects;
    drafting “benchmark roadmaps,” which identify categories of benchmarks that could be valuable in the future and outline prerequisites for developing them.
    And AI governance professionals can help by:
    co-organizing workshops, competitions, and prizes focused on benchmarking;
    creating third-party institutional homes for benchmarking work;
    clarifying, ahead of time, how auditing and regulatory frameworks can put benchmarks to use;
    advising safety researchers on political, institutional, and strategic considerations that matter for benchmark design;
    popularizing the narrative of a “race to the top” on AI safety.
    Ultimately, we can and should begin to build benchmark-making capability now.
    Acknowledgment
    I would like to thank Ben Garfinkel and Owen Cotton-Barratt for their mentorship, Emma Bluemke and many others at the Centre for the Governance of AI for their warmhearted support. All views and errors are my own.
    Future research
    I am working on a paper on the topic, and if you are interested in benchmarks and model evaluation, especially if you are a technical AI safety researcher, I would love to hear from you!
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    3 min
  • LW - Probably good projects for the AI safety ecosystem by Ryan Kidd
    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: Probably good projects for the AI safety ecosystem, published by Ryan Kidd on December 5, 2022 on LessWrong.
    At EAGxBerkeley 2022, I was asked several times what new projects might benefit the AI safety and longtermist research ecosystem. I think that several existing useful-according-to-me projects (e.g., SERI MATS, REMIX, CAIS, etc.) could urgently absorb strong management and operations talent, but I think the following projects would also probably be useful to the AI safety/longtermist project. Criticisms are welcome.
    Projects I might be excited to see, in no particular order:
    A London-based MATS clone to build the AI safety research ecosystem there, leverage mentors in and around London (e.g., DeepMind, CLR, David Krueger, Aligned AI, Conjecture, etc.), and allow regional specialization. This project should probably only happen once MATS has ironed out the bugs in its beta versions and grown too large for one location (possibly by Winter 2023). Please contact the MATS team before starting something like this to ensure good coordination and to learn from our mistakes.
    Rolling admissions alternatives to MATS’ cohort-based structure for mentors and scholars with different needs (e.g., to support alignment researchers who suddenly want to train/use research talent at irregular intervals but don’t have the operational support to do this optimally).
    A combined research mentorship and seminar program that aims to do for AI governance research what MATS is trying to do for technical AI alignment research.
    A dedicated bi-yearly workshop for AI safety university group leaders that teaches them how to recognize talent, foster useful undergraduate research projects, and build a good talent development pipeline or “user journey” (including a model of alignment macrostrategy and where university groups fit in).
    An organization that does for the Open Philanthropy worldview investigations team what GCP did to supplement CEA's workshops and 80,000 Hours’ career advising calls.
    Further programs like ARENA that aim to develop ML safety engineering talent at scale by leveraging good ML tutors and proven curricula like CAIS’ Intro to ML Safety, Redwood Research’s MLAB, and Jacob Hilton's DL curriculum for large language module alignment.
    More contests like ELK with well-operationalized research problems (i.e., clearly explain what builder/breaker steps look like), clear metrics of success, and have a well-considered target audience (who is being incentivized to apply and why?) and user journey (where do prize winners go next?). Possible contest seeds:
    Evan Hubinger’s SERI MATS deceptive AI challenge problem;
    Vivek Hebbar’s and Nate Soares’ SERI MATS diamond maximizer selection problem;
    Alex Turner’s and Quintin Pope’s SERI MATS training stories selection problem.
    More "plug-and-play" curriculums for AI safety university groups, like AGI Safety Fundamentals, Alignment 201, Intro to ML Safety.
    A well-considered "precipism" university course template that critically analyzes Toby Ord's “The Precipice,” Holden Karnofsky's “The Most Important Century,” Will MacAskill's “What We Owe The Future,” some Open Philanthropy worldview investigations reports, some Global Priorities Institute ethics papers, etc.
    Hackathons in which people with strong ML knowledge (not ML novices) write good-faith critiques of AI alignment papers and worldviews (e.g., what Jacob Steinhardt’s “ML Systems Will Have Weird Failure Modes” does for Hubinger et al.’s “Risks From Learned Optimization”).
    A New York-based alignment hub that aims to provide talent search and logistical support for NYU Professor Sam Bowman’s planned AI safety research group.
    More organizations like CAIS that aim to recruit established ML talent into alignment research with clear benchmarks, targeted hackathons/contests with prizes, and offers ...
    5 min
  • EA - EA Taskmaster Game by Gemma Paterson
    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 Taskmaster Game, published by Gemma Paterson on December 4, 2022 on The Effective Altruism Forum.
    Introduction
    This is a post about a silly EA themed game I organised in London this year. All resources included in this post are free for you to use for your own EA events with the following caveats:
    If after reading this you have task ideas, please submit them here and they might be used in my next event: GoogleForm | EA Taskmaster 2023 Task Ideas
    If you want to run one of these yourself, I recommend reading my thoughts and suggestions in the logistical points for organisers section
    If you do run one of these, I would love to hear about it (and see photos ). Please get in touch - you can message me here on the forum or contact me on Twitter glpat (@glpat99) / Twitter.
    I had a fabulous time putting it together and the feedback from attendees was really positive so if this sounds like your kind of “organised fun” then I hope this post is helpful/inspiring. Thanks to Romy and Vania from the London EA women & NB meet-up who pushed me to make this my first forum post!
    The Game
    In June 2022, I ran an EA themed game that was a mix between a scavenger hunt, the TV show Taskmaster and a party quest game. It was made up of 89 tasks (some EA themed, some just silly) to be completed by competing teams over the course of an afternoon.
    The idea was that there would be too many tasks for a team to complete in the time limit and they would have to do some task prioritisation. When I was thinking about what tasks should be included, I wanted a range of tasks that could be thought of in terms of scale, neglectedness and tractability, so that it would be a much lower stakes game version of cause prioritisation:
    Scale: There was a large variation in point values across the tasks, not necessarily in proportion to their difficulty. Also, some tasks were repeatable (with accumulating points) while others were one-offs.
    Neglectedness: For some tasks the points would only go to the team with the best entry or the points would be split between all teams that submitted. This incentivised completing tasks that no one else was completing.
    Tractability: There was a large range of task difficulty, with hard tasks not necessarily being rewarded with extra points.
    In fact, one of the tasks was to send in the best task prioritisation spreadsheet for the tasks in this game. As per the Taskmaster game show, final point allocation was at the discretion of the Taskmaster (me) and was based on ingenuity, grit and entertainment value.
    Huge thanks to my EA friends in London for helping me think of tasks. Also, shout out to the Party Quest spreadsheet for providing many of the fun non-EA themed filler tasks.
    I am conscious that posting all the tasks would spoil the game for potential players so I have provided a small selection of some of the tasks and submissions from this year’s event.
    Task examples
    Task 24: Sculptor (2000 points)
    Construct the most beautiful trophy out of random objects and deliver it to the Taskmaster.
    The winner’s podium with the constructed trophies given in order of the Taskmaster’s preference
    Task 19: I’m Thirsty (800 points)
    Bring the taskmaster the best cocktail.
    The Taskmaster on her homemade throne graciously accepting her third cocktail of the afternoon
    Task 51: An Order From the Ministry (500 points)
    Create the silliest walk
    Less Miserables' excellent submission
    Task 67: Very Culture, Much Topical, Wow (1000 points)
    Recreate a meme in real life and photograph/video it. The more elaborate the better.
    Task 68: Getting down with the kids (1000 points)
    Create an EA themed tiktok
    Pink team's novel new approach to AI safety
    Structure of game
    There were four categories of task which were based on how points are allocated.
    Winner take all: All the points were awarded to the team whose...
    11 min
  • EA - Applications for EAGx LatAm closing on the 20th of December by LGlez
    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: Applications for EAGx LatAm closing on the 20th of December, published by LGlez on December 4, 2022 on The Effective Altruism Forum.
    This is just a short announcement that you can now apply to EAGx LatAm, which will take place in Mexico City on January 6-8. For more information, you can read our previous announcement here. Apologies for the delay in getting this post out, but we had to postpone it for circumstances beyond our control.
    “How to do the most good” is a very hard problem and no isolated community will solve it on its own. We are excited about opening spaces to discuss this question in a variety of contexts, with diversity and inclusion in mind. The unfortunate fact is that it’s easier for some to hear about EA and be heard in EA, for reasons that have nothing to do with talent on one side or ill intent on the other, and everything to do with visas, the language you happened to be raised with, your socioeconomic background or environment. We believe that EA’s current blindspots can be identified by brilliant minds all over the world, and we want to promote spaces for more people to come together, learn and bring their unique perspectives to the conversation.
    We have almost finalised the programme. You can expect to see sessions to discuss community building in Low and Middle Income Countries and meet organisations working on top EA areas outside of the UK and the US. We will also have the usual EAGx suspects, including intro talks about Global Health and Development, AI Safety, Animal Welfare or Global Catastrophic Risks. We have confirmed speakers from GWWC, CEA, Charity Entrepreneurship, JPAL, Open Philanthropy, IPA, Rethink Priorities and the World Bank, among many others. Rob Wiblin will come to practice his Spanish talk about Career Prioritisation and Toby Ord will join us virtually to discuss longtermism in the context of the Global South.
    This conference is mainly for those who are from Latin America or have ties to the continent, because they are the ones for whom it’s most difficult to attend events elsewhere and connect with people who face similar struggles. But as we were saying before, we don’t believe in isolated communities and we certainly wouldn’t want to build a segregated EA LatAm community! On the contrary, we see this as an opportunity for connection. We’re therefore looking forward to receiving experienced members of the international community who are excited to meet talented people who might be under their radar –to talk with them, mentor them, learn from them, collaborate with them, hire them. On the other hand, if you’re new to EA, willing to learn more and based in a region that hosts regular EA conferences (e.g. the US, Europe), we suggest one of those might be a better fit for you to be introduced to the movement.
    If in doubt, err on the side of applying.
    And no, you don’t need to speak Spanish to join us.
    If you’re asking for funding to cover your travel expenses, bear in mind that costs in Mexico are cheap compared to Europe and the US. You should be able to find a good hotel for ∼60$. Regarding transport, Uber is available and very cheap.
    If you have any questions or comments don’t hesitate to write to [email protected]
    We think that EA with a Latin American spice can be great. Join us in January to discuss doing good over tacos and get the coolest EA merch the world has ever seen.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    4 min