The Nonlinear Library

The Nonlinear Library

By The Nonlinear FundEducation
Download on the App Store

All The Nonlinear Library episodes

  • EA - Shaping Humanity's Longterm Trajectory by Toby Ord
    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: Shaping Humanity's Longterm Trajectory, published by Toby Ord on July 18, 2023 on The Effective Altruism Forum.
    Since writing The Precipice, one of my aims has been to better understand how reducing existential risk compares with other ways of influencing the longterm future. Helping avert a catastrophe can have profound value due to the way that the short-run effects of our actions can have a systematic influence on the long-run future. But it isn't the only way that could happen.
    For example, if we advanced human progress by a year, perhaps we should expect to see us reach each subsequent milestone a year earlier. And if things are generally becoming better over time, then this may make all years across the whole future better on average.
    I've developed a clean mathematical framework in which possibilities like this can be made precise, the assumptions behind them can be clearly stated, and their value can be compared.
    The starting point is the longterm trajectory of humanity, understood as how the instantaneous value of humanity unfolds over time. In this framework, the value of our future is equal to the area under this curve and the value of altering our trajectory is equal to the area between the original curve and the altered curve.
    This allows us to compare the value of reducing existential risk to other ways our actions might improve the longterm future, such as improving the values that guide humanity, or advancing progress.
    Ultimately, I draw out and name 4 idealised ways our short-term actions could change the longterm trajectory:
    advancements
    speed-ups
    gains
    enhancements
    And I show how these compare to each other, and to reducing existential risk.
    While the framework is mathematical, the maths in these four cases turns out to simplify dramatically, so anyone should be able to follow it.
    My hope is that this framework, and this categorisation of some of the key ways we might hope to shape the longterm future, can improve our thinking about longtermism.
    Some upshots of the work:
    Some ways of altering our trajectory only scale with humanity's duration or its average value - but not both. There is a serious advantage to those that scale with both: speed-ups, enhancements, and reducing existential risk.
    When people talk about 'speed-ups', they are often conflating two different concepts. I disentangle these into advancements and speed-ups, showing that we mainly have advancements in mind, but that true speed-ups may yet be possible.
    The value of advancements and speed-ups depends crucially on whether they also bring forward the end of humanity. When they do, they have negative value.
    It is hard for pure advancements to compete with reducing existential risk as their value turns out not to scale with the duration of humanity's future. Advancements are competitive in outcomes where value increases exponentially up until the end time, but this isn't likely over the very long run. Work on creating longterm value via advancing progress is most likely to compete with reducing risk if the focus is on increasing the relative progress of some areas over others, in order to make a more radical change to the trajectory.
    The work is appearing as a chapter for the forthcoming book, Essays on Longtermism, but as of today, you can also read it online here.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org
    4 min
  • EA - New career review: AI safety technical research by Benjamin Hilton
    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 career review: AI safety technical research, published by Benjamin Hilton on July 18, 2023 on The Effective Altruism Forum.
    Note: this post is a (minorly) edited version of a new 80,000 Hours career review.
    Progress in AI - while it could be hugely beneficial - comes with significant risks. Risks that we've argued could be existential.
    But these risks can be tackled.
    With further progress in AI safety, we have an opportunity to develop AI for good: systems that are safe, ethical, and beneficial for everyone.
    This article explains how you can help.
    Summary
    Artificial intelligence will have transformative effects on society over the coming decades, and could bring huge benefits - but we also think there's a substantial risk. One promising way to reduce the chances of an AI-related catastrophe is to find technical solutions that could allow us to prevent AI systems from carrying out dangerous behaviour.
    Pros
    Opportunity to make a significant contribution to a hugely important area of research
    Intellectually challenging and interesting work
    The area has a strong need for skilled researchers and engineers, and is highly neglected overall
    Cons
    Due to a shortage of managers, it's difficult to get jobs and might take you some time to build the required career capital and expertise
    You need a strong quantitative background
    It might be very difficult to find solutions
    There's a real risk of doing harm
    Key facts on fit
    You'll need a quantitative background and should probably enjoy programming. If you've never tried programming, you may be a good fit if you can break problems down into logical parts, generate and test hypotheses, possess a willingness to try out many different solutions, and have high attention to detail.
    If you already:
    Are a strong software engineer, you could apply for empirical research contributor roles right now (even if you don't have a machine learning background, although that helps)
    Could get into a top 10 machine learning PhD, that would put you on track to become a research lead
    Have a very strong maths or theoretical computer science background, you'll probably be a good fit for theoretical alignment research
    Recommended
    If you are well suited to this career, it may be the best way for you to have a social impact.
    Thanks to Adam Gleave, Jacob Hilton and Rohin Shah for reviewing this article. And thanks to Charlie Rogers-Smith for his help, and his article on the topic - How to pursue a career in technical AI alignment.
    Why AI safety technical research is high impact
    As we've argued, in the next few decades, we might see the development of hugely powerful machine learning systems with the potential to transform society. This transformation could bring huge benefits - but only if we avoid the risks.
    We think that the worst-case risks from AI systems arise in large part because AI systems could be misaligned - that is, they will aim to do things that we don't want them to do. In particular, we think they could be misaligned in such a way that they develop (and execute) plans that pose risks to humanity's ability to influence the world, even when we don't want that influence to be lost.
    We think this means that these future systems pose an existential threat to civilisation.
    Even if we find a way to avoid this power-seeking behaviour, there are still substantial risks - such as misuse by governments or other actors - which could be existential threats in themselves.
    There are many ways in which we could go about reducing the risks that these systems might pose. But one of the most promising may be researching technical solutions that prevent unwanted behaviour - including misaligned behaviour - from AI systems. (Finding a technical way to prevent misalignment in particular is known as the alignment problem.)
    In the past few years, we've seen more o...
    48 min
  • EA - Effective Altruism and the strategic ambiguity of 'doing good' by Jeroen De Ryck
    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 and the strategic ambiguity of 'doing good', published by Jeroen De Ryck on July 18, 2023 on The Effective Altruism Forum.
    Whilst Googling around for something entirely unrelated, I stumbled on a discussion paper published in January of 2023 about Effective Altruism that argues Global Health & Wellbeing is basically a facade to get people into the way more controversial core of longtermism. I couldn't find something posted about it elsewhere on the forum, so I'll try to summarise here.
    The paper argues that there is a big distinction between what they call public facing EA and Core EA. The former cares about global health and wellbeing (GH&W) whereas the latter cares about x-risks, animal welfare and "helping elites get advanced degrees" (which I'll just refer to as core topics). There are several more distinctions between public EA and core EA, e.g. about impartiality and the importance of evidence and reason. The author argues, based on quotes from a variety of posts from a variety of influential people within EA, that for the core audience, GH&W is just a facade such that EA is perceived as 'good' by the broader public, whilst the core members work on much more controversial core topics such as transhumanism that go against many of the principles put forward by GH&W research and positions.
    The author seems to claim that this was done on purpose and that GH&W merely exists as a method to "convert more recruits" to a controversial core of transhumanism that EA is nowadays. This substantial distinction between GH&W and core topics causes an identity crisis between people who genuinely believe that EA is about GH&W and people who have been convinced of the core topics. The author says that these distinctions have always existed, but have been purposely hidden with nice-sounding GH&W topics by a few core members (such as Yudkowsky, Alexander, Todd, Ord, MacAskill), as a transhumanist agenda would be too controversial for the public, although it was the goal of EA after all and always has been.
    To quote from the final paragraph from the paper:
    The 'EA' that academics write about is a mirage, albeit one invoked as shorthand for a very real phenomenon, i.e., the elevation of RCTs and quantitative evaluation methods in the aid and development sector. [...] Rather, my point is that these articles and the arguments they make - sophisticated and valuable as they are - are not about EA: they are about the Singer-solution to global poverty, effective giving, and about the role of RCTs and quantitative evaluation methods in development practice. EA is an entirely different project, and the magnitude and implications of that project cannot be grasped until people are willing to look at the evidence beyond EA's glossy front-cover, and see what activities and aims the EA movement actually prioritizes, how funding is actually distributed, whose agenda is actually pursued, and whose interests are actually served.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org
    3 min
  • LW - Existential Risk Persuasion Tournament by PeterMcCluskey
    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: Existential Risk Persuasion Tournament, published by PeterMcCluskey on July 17, 2023 on LessWrong.
    I participated last summer in Tetlock's Existential Risk Persuasion
    Tournament (755(!) page paper
    here).
    Superforecasters and "subject matter experts" engaged in a hybrid between a prediction market and debates, to predict catastrophic and existential risks this century.
    I signed up as a superforecaster. My impression was that I knew as much about AI risk as any of the subject matter experts with whom I interacted (the tournament was divided up so that I was only aware of a small fraction of the 169 participants).
    I didn't notice anyone with substantial expertise in machine learning. Experts were apparently chosen based on having some sort of respectable publication related to AI, nuclear, climate, or biological catastrophic risks. Those experts were more competent, in one of those fields, than news media pundits or politicians. I.e. they're likely to be more accurate than random guesses. But maybe not by a large margin.
    That expertise leaves much to be desired. I'm unsure whether there was a realistic way for the sponsors to attract better experts. There seems to be not enough money or prestige to attract the very best experts.
    Incentives
    The success of the superforecasting approach depends heavily on forecasters having decent incentives.
    It's tricky to give people incentives to forecast events that will be evaluated in 2100, or evaluated after humans go extinct.
    The tournament provided a fairly standard scoring rule for questions that resolve by 2030. That's a fairly safe way to get parts of the tournament to work well.
    The other questions were scored by how well the forecast matched the median forecast of other participants (excluding participants that the forecasters interacted with). It's hard to tell whether that incentive helped or hurt the accuracy of the forecasts. It's easy to imagine that it discouraged forecasters from relying on evidence that is hard to articulate, or hard to verify. It provided an incentive for groupthink. But the overall incentives were weak enough that altruistic pursuit of accuracy might have prevailed. Or ideological dogmatism might have prevailed. It will take time before we have even weak evidence as to which was the case.
    One incentive that occurred to me toward the end of the tournament was the possibility of getting a verified longterm forecasting track record. Suppose that in 2050 they redo the scores based on evidence available then, and I score in the top 10% of tournament participants. That would likely mean that I'm one of maybe a dozen people in the world with a good track record for forecasting 28 years into the future. I can imagine that being valuable enough for someone to revive me from cryonic suspension when I'd otherwise be forgotten.
    There were some sort of rewards for writing comments that influenced other participants. I didn't pay much attention to those.
    Quality of the Questions
    There were many questions loosely related to AGI timelines, none of them quite satisfying my desire for something closely related to extinction risk that could be scored before it's too late to avoid the risk.
    One question was based on a Metaculus forecast for an advanced AI. It seems to represent clear progress toward the kind of AGI that could cause dramatic changes. But I expect important disagreements over how much progress it represents: what scale should we use to decide how close such an AI is to a dangerous AI? does the Turing test use judges who have expertise in finding the AI's weaknesses?
    Another question was about when Nick Bostrom will decide that an AGI exists. Or if he doesn't say anything clear, then a panel of experts will guess what Bostrom would say. That's pretty close to a good question to forecast. Can we assume tha...
    14 min
  • LW - Predictive history classes by dkl9
    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: Predictive history classes, published by dkl9 on July 17, 2023 on LessWrong.
    Why do we study history? There are many potential reasons. More specifically, why do we teach history to everyone going to school? Many reasons become less relevant to those other than historians or history teachers.
    The classic reason is "so that we don't repeat it". But most people will not end up in a position of power so as to decide whether we repeat or don't repeat history. A more honest reason, for most people, is that we learn history to generalise its lessons to the future or present. Assuming that purpose, we find current methods of history classes dreadfully ineffective. I propose an alternative paradigm of history classes which, I believe, will result in much more practical learning.
    A cynical reader remarks that history education in school (in the form I'm targeting) is for indoctrination. From that view, take this not as a proposal for reform, but a suggestion on how to study well for your own sake.
    Current history education focuses on teaching the students about past events. Students are then tested on those same past events. In higher history classes, they are also tested on analysing information about these events and arguing about their causes and consequences.
    This is great for those who will be historians. Most students will not. This style of teaching does not guarantee any understanding of how to apply historical lessons to the present and future.
    In schools where students just study to the test, make the test a good one. Instead of testing students on past events they studied, test them on events they didn't study. They are then forced to learn how to generalise history, applying its lessons to understand what they're going thru and what may come next. To get an objective answer by which to grade the students, test them on actual past events - just obscure ones that they wouldn't already know about. Sith they didn't study the material, they aren't expected to ever get a perfect score. This would make history classes more like maths or science classes, in that the students learn methods to creatively apply to problems they haven't exactly seen before.
    This guiding principle should apply beyond the tests. We would replace history classes with predictive history classes. We teach about events and the patterns which underlie them, assign students to study and predict parts of new-to-them events on their own, and test by soliciting predictions on unfamiliar bits of history.
    The argument for predictive history classes comes not just from mere pragmatism. Epistemically, current history classes are almost devoid of content. We teach students to look at events and explain them, but they never have to question what they're told. Sure, they're to think critically about conclusions from the events or claims regarding their causes, but the event itself goes unquestioned. They should be able to recognise which events are plausible or implausible iff they have a true understanding of history. Current history courses make no effort to teach them this. If you are equally good at explaining any outcome, you have no knowledge.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    3 min
  • EA - What do y'all think of pay gaps in EA? by JohnSnow
    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 do y'all think of pay gaps in EA?, published by JohnSnow on July 17, 2023 on The Effective Altruism Forum.
    I saw a CEA role advertised for twice the salary of an AMF job, whereas they do not seem to differ dramatically in expected level of expertise / experience. Even if one believes they can make more impact at AMF, they would have to give up 20k pounds in salary to pass on the content specialist role. We learned recently to consider earning less, but this may still be quite the conundrum. What do you think?
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org
    1 min
  • LW - AutoInterpretation Finds Sparse Coding Beats Alternatives 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: AutoInterpretation Finds Sparse Coding Beats Alternatives, published by Hoagy on July 17, 2023 on LessWrong.
    Produced as part of the SERI ML Alignment Theory Scholars Program - Summer 2023 Cohort
    Huge thanks to Logan Riggs, Aidan Ewart, Lee Sharkey, Robert Huben for their work on the sparse coding project, Lee Sharkey and Chris Mathwin for comments on the draft, EleutherAI for compute and OpenAI for GPT-4 credits.
    Summary
    We use OpenAI's automatic interpretation protocol to analyse features found by dictionary learning using sparse coding and compare the interpretability scores thereby found to a variety of baselines.
    We find that for both the residual stream (layer 2) and MLP (layer 1) of Eleuther's Pythia70M, sparse coding learns a set of features that is superior to all tested baselines, even when removing the bias and looking just at the learnt directions. In doing so we provide additional evidence to the hypothesis that NNs should be conceived as using distributed representations to represent linear features which are only weakly anchored to the neuron basis.
    As before these results are still somewhat preliminary and we hope to expand on them and make them more robust over the coming month or two, but we hope people find them fruitful sources of ideas. If you want to discuss, feel free to message me or head over to our thread in the EleutherAI discord.
    All code available at the github repo.
    Methods
    Sparse Coding
    The feature dictionaries learned by sparse coding are learnt by simple linear autoencoders with a sparsity penalty on the activations. For more background on the sparse coding approach to feature-finding see the Conjecture interim report that we're building from, or Robert Huben's explainer.
    Automatic Interpretation
    As Logan Riggs' recently found, many of the directions found through sparse coding seem highly interpretable, but we wanted a way to quantify this, and make sure that we were detecting a real difference in the level of interpretability.
    To do this we used the methodology outlined in this OpenAI paper, details can be found in their code base. To quickly summarise, we are analysing features which are defined as scalar-valued functions of the activations of a neural network, limiting ourselves here to features defined on a single layer of a language model. The original paper simply defined features as the activation of individual neurons but we will in general be looking at linear combinations of neurons.
    We give a feature an interpretability score by first generating a natural language explanation for the feature, which is expected to explain how strongly a feature will be active in a certain context, for example 'the feature activates on legal terminology'. Then, we give this explanation to an LLM and ask it to predict the feature for hundreds of different contexts, so if the tokens are ['the' 'lawyer' 'went' 'to' 'the' 'court'] the predicted activations might be [0, 10, 0, 0, 8]. The score is defined as the correlation between the true and predicted activations.
    To generate the explanations we follow OpenAI and take a 64-token sentence-fragment from each of the first 50,000 lines of OpenWebText. For each feature, we calculate the average activation and take the 20 fragments with the highest activation. Of these 20, we pass 5 to GPT-4, along with the rescaled per-token activations. From these 5 fragments, GPT-4 suggests an explanation for when the neuron fires.
    GPT3.5 is then used to simulate the feature, given the explanation, across both another 5 highly activating fragments, and 5 randomly selected fragments (with non-zero variation). The correlation scores are calculated across all 10 fragments ('top-and-random'), as well as for the top and random fragments separately.
    Comparing Feature Dictionaries
    We use dictionary learning with a sparsi...
    12 min
  • LW - Sapient Algorithms by Valentine
    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: Sapient Algorithms, published by Valentine on July 17, 2023 on LessWrong.
    I notice my mind runs lots of cached programs. Like "walk", "put away the dishes", "drive home", "go to the bathroom", "check phone", etc.
    Most of these can run "on autopilot". I don't know how to define that formally. But I'm talking about how, e.g., I can start driving and get lost in thought and suddenly discover I'm back home - sometimes even if that wasn't where I was trying to go!
    But some programs cannot run on autopilot. The algorithm has something like a "summon sapience" step in it. Even if the algorithm got activated due to autopilot, some step turns it off.
    When I look at the examples of sapient algorithms that I run, I notice they have a neat kind of auto-generalization nature to them. I have some reason to think that property is general. It's the opposite of how, e.g., setting up webpage blockers can cause my fingers to autopilot learn how to bypass them.
    I'll try to illustrate what I mean via examples.
    Example: Look at my car keys
    I got tired of risking locking my keys in my car. So I started making a habit of looking at my keys before closing the door.
    Once, right after I'd closed the locked car door, I realized I'd looked at the phone in my hand and shut the door anyway. Luckily the key was in my pocket. But I noticed that this autopilot program just wasn't helping.
    So I modified it (as a TAP): If I was about to close the car door, I would look at my hand, turn on consciousness, and check if I was actually looking at my keys.
    First, that TAP just worked. To this day I still do this when stepping out of a car.
    Second, it generalized without my trying to:
    After a while it would fire whenever I was about to close any locked door.
    It then generalized to anyone I was with. If they were about to close a locked door, I would sort of "pop awake" with a mental question about whether someone had the key.
    It then generalized even more. It now fires when I'm, say, preparing for international travel. Crossing a border feels a bit like going through a door that locks behind me. So now I "wake up" and check that I and my travel companions all have passports. (I would usually check before, but now it's specifically this mental algorithm that summons sapience. It's reliable instead of being an extra thing to remember.)
    This generalization wasn't intentional. But it's been really good. I haven't noticed any problems at all from this program sort of spreading on its own.
    Example: Taste my food
    When I'm in autopilot mode while eating, it can feel at the end like my food kind of vanished. Like I wasn't there for the meal.
    So I installed a TAP: If I'm about to put food in my mouth, pause & remember emptiness.
    "Remember emptiness" has a "summon sapience" type move embedded in it. It's something like "Turn on consciousness, pause, and really look at my sensory input." It's quite a bit deeper than that, but if this kind of emptiness just sounds like gobbledegook to you then you can pretend I said the simplified version.
    In this case, the TAP itself didn't install as cleanly as with the car keys example. Sometimes I just forget. Sometimes the TAP fires only after I've taken my first bite.
    But all the same, the algorithm still sort of auto-generalized. When I'm viewing a beautiful vista, or am part of a touching conversation, or hear some lovely music, the TAP sometimes fires (about as regularly as with food). One moment there are standard programs running, and then all of a sudden "I'm there" and am actually being touched by whatever it is (the same way I'm actually tasting my food when I'm "there").
    Yesterday I noticed this sapient algorithm booting up in a conversation. Someone asked me "Can I speak plainly?" and I knew she was about to say something I'd find challenging to receive. My autopilot start...
    8 min
  • LW - Thoughts on "Process-Based Supervision" by Steven Byrnes
    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: Thoughts on "Process-Based Supervision", published by Steven Byrnes on July 17, 2023 on LessWrong.
    1. Post Summary / Table of Contents
    In "How might we align transformative AI if it's developed very soon?", Holden Karnofsky talked about "Process-Based Supervision", citing a previous post by Stuhlmüllert & Byun of Ought. (Holden says he got the idea mainly from Paul Christiano.)
    I apparently misunderstood what Holden meant by "Process-Based Supervision", and it took many hours and a 7000-word comment thread before I figured it out.
    (Thanks to Holden for his extraordinary patience during that protracted discussion.)
    The extremely short version for AI alignment domain experts is: I currently think of the load-bearing ingredients of Holden's take on "process-based supervision" as being:
    AI boxing some of the time (specifically, during the periods where the AI is "thinking");
    "Myopic training" (e.g. as defined here);
    NOT aspiring to be a complete solution to safety, but rather a more modest attempt to help avoid situations where we (accidentally) positively reinforce the AI for engaging in directly dangerous behavior.
    You could think of this as an intervention that directly and straightforwardly mitigates outer misalignment, and that's the main thing I'll discuss in this post. But obviously, any change of the supervisory signals will have some effect on the likelihood of inner misalignment / goal misgeneralization too. And there's also a (more speculative) argument that process-based supervision might make things better there too - at least on the margin. See footnote 4 of Section 5.2.2.
    (This is specific to Holden's take. I think Stuhlmüllert & Byun's take on "process-based supervision" involves a different set of load-bearing ingredients, centered around restricting the complexity of black-box processing. I will not be discussing that.)
    The long, hopefully-pedagogical, and more opinionated version is the rest of this post.
    Table of Contents:
    Section 2 will give the very brief slogan / sales-pitch for process-based supervision, and why that pitch was bouncing off me, striking me as frustratingly missing-the-point.
    Section 3 will state the subproblem that we're trying to solve: the AI does subtly manipulative, power-seeking, or otherwise problematic actions, and we don't notice, and therefore we give a training signal that reinforces that behavior, and therefore the AI does those things more and more. To be clear, this is not the only path to dangerous misalignment (in particular, classic "treacherous turns" are out-of-scope). But maybe solving just this subproblem can be part of a complete solution. I'll get back to that in Section 5.
    Section 4 describes "process-based supervision" as I currently understand it, and why it seems to solve the subproblem in question.
    Finally, having described process-based supervision as I currently understand it, Section 5 offers a critical evaluation of that idea. In particular:
    5.1 asks "Does this actually solve the subproblem in question?";
    5.2 asks "What about the other misalignment-related subproblems?";
    5.3 asks "How bad is the "alignment tax" from doing this kind of thing?";
    and 5.4 is a summary.
    Tl;dr: Once we get to the capabilities regime where AI safety / alignment really matters, I currently think that process-based supervision would entail paying a very big alignment tax - actually, not just "big" but potentially infinite, as in "this kind of AGI just plain can't do anything of significance". And I also currently think that, of the somewhat-vague paths I see towards AGI technical safety, process-based supervision wouldn't make those paths noticeably easier or more likely to succeed. (Of those two complaints, I feel more strongly about the first one.) This take is pretty specific to my models of what AGI algorithms will look li...
    42 min
  • EA - EA Organization Updates: July 2023 by Lizka
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: EA Organization Updates: July 2023, published by Lizka on July 17, 2023 on The Effective Altruism Forum.
    We're featuring some opportunities and job listings at the top of this post. Some have (very) pressing deadlines.
    You can see previous updates on the "EA Organization Updates (monthly series)" topic page, or in our repository of past newsletters. Notice that there's also an "org update" tag, where you can find more news and updates that are not part of this consolidated series.
    These monthly posts originated as the "Updates" section of the EA Newsletter. Organizations submit their own updates, which we edit for clarity.
    (If you think your organization should be getting emails about adding their updates to this series, please apply here.)
    Opportunities and jobs
    Opportunities
    Consider also checking opportunities listed on the EA Opportunities Board.
    Applications are open for a number of conferences
    A number of Effective Altruism Global and EAGx conferences have upcoming deadlines:
    EAGxNYC will run 18-20 August. Tickets are $0-100. If you live in the New York area, consider applying by 31 July.
    EAGxBerlin, runs 8-10 September and is aimed at people in Western Europe. Tickets cost €0-80. Apply by 18 August.
    EAGxAustralia (22 - 24 September) is for people in Australia and New Zealand. Tickets are $75-150 (AUD). Apply by 8 September.
    Other conferences with open applications include EA Global: Boston (27-29 October, apply by 13 October), and EAGxPhilippines.
    The international Conference on Animal Rights in Europe (CARE) will run 17-20 August, in Warsaw and online. Participants from all areas of expertise are invited to network with activists, discover funding opportunities, and build a stronger movement for animals. You can get CARE 2023 tickets until 1 August.
    Fellowships and incubation programs
    GovAI's 2024 Winter Fellowship gives people the opportunity to spend three months (February-April) working on an AI governance project, learning about the field, and networking. Fellows get a £9,000 stipend and support for traveling to Oxford. If you're early in your career and are interested in studying or shaping the long-term implications of AI, consider applying by 23 July.
    Charity Entrepreneurship is accepting applications for its charity incubation programs in 2024 (apply by 30 September) and for its new online Research Training Program (2 October-17 December - apply by 17 July - today). The research program focuses on tools and skills needed to identify, compare, and recommend the most cost-effective and evidence-based charities and interventions. It is a full-time, fully cost-covered program that will run online for 11 weeks.
    Other opportunities
    The Roots of Progress Blog Building Intensive is for writers eager to write more (and better) about progress studies topics. Participants will connect with other writers, receive writing coaching, and more. The part-time, 8-week online program runs from mid-September to mid-November. Apply by 11 August.
    There's a new regranting platform, Manifund; you can apply for grants, apply to regrant, or just explore and participate in the discussion.
    If you're interested in running an EA conference in your region or country (EAGx), you can apply for support.
    Ian Hogarth, the new Chair of the UK's AI Foundation Model Taskforce invites expressions of interest from AI specialists who want to help.
    Job listings
    Consider also exploring jobs listed on "Job listing (open)."
    Against Malaria Foundation
    Junior Operations Manager (Remote, £28K - £35K)
    Centre for Effective Altruism
    Content Specialist (Remote / Oxford / Boston / other, £54.6k -£67.3k / $96.2-$124.k, apply by 26 July)
    Cooperative AI Foundation
    Managing Director (Remote, $100K - $130K, apply by 30 July)
    GiveWell
    Senior Researcher (Remote or Oakland, CA; $193,100 - $209,000)
    Content Edito...
    18 min