The Nonlinear Library

The Nonlinear Library

By The Nonlinear FundEducation
Download on the App Store

All The Nonlinear Library episodes

  • AF - Short Remark on the (subjective) mathematical 'naturalness' of the Nanda--Lieberum addition modulo 113 algorithm by Spencer Becker-Kahn
    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: Short Remark on the (subjective) mathematical 'naturalness' of the Nanda--Lieberum addition modulo 113 algorithm, published by Spencer Becker-Kahn on June 1, 2023 on The AI Alignment Forum.
    These remarks are basically me just wanting to get my thoughts down after a Twitter exchange on this subject. I've not spent much time on this post and it's certainly plausible that I've gotten things wrong.In the 'Key Takeaways' section of the Modular Addition part of the well-known post 'A Mechanistic Interpretability Analysis of Grokking' , Nanda and Lieberum write:
    This algorithm operates via using trig identities and Discrete Fourier Transforms to map x,ycos(w(x+y)),sin(w(x+y)), and then extracting x+y(modp)
    And
    The model is trained to map x,y to z≡x+y(mod113) (henceforth 113 is referred to as p)
    But the casual reader should use caution! It is in fact the case that "Inputs x,y are given as one-hot encoded vectors in Rp ". This point is of course emphasized more in the full notebook (it has to be, that's where the code is), and the arXiv paper that followed is also much clearer about this point. However, when giving brief takeaways from the work, especially when it comes to discussing how 'natural' the learned algorithm is, I would go as far as saying that it is actually misleading to suggest that the network is literally given x and y as inputs. It is not trained to 'act' on the numbers x, y themselves. When thinking seriously about why the network is doing the particular thing that it is doing at the mechanistic level, I would want to emphasize that one-hotting is already a significant transformation. You have moved away from having the number x be represented by its own magnitude.
    You instead have a situation in which x and y now really live 'in the domain' (its almost like a dual point of view: The number x is not the size of the signal, but the position at which the input signal is non-zero). So, while I of course fully admit that I too am looking at it through my own subjective lens, one might say that (before the embedding happens) it is more mathematically natural to think that what the network is 'seeing' as input is something like the indicator functions t↦1x(t) and t↦1y(t). Here, t is something like the 'token variable' in the sense that these are functions on the vocabulary. And if we essentially ignore the additional tokens for | and =, we can think that these are functions on the group Z/pZ and that we would like the network to learn to produce the function t↦1x+y(t) at its output neurons.In particular, this point of view further (and perhaps almost completely) demystifies the use of the Fourier basis.
    Notice that the operation you want to learn is manifestly a convolution operation, i.e.
    And (as I distinctly remember being made to practically chant in an 'Analysis of Boolean Functions' class given by Tom Sanders) the Fourier Transform is the (essentially unique) change of basis that simultaneously diagonalizes all convolution operations. This is coming close to saying something like: There is one special basis that makes the operation you want to learn uniquely easy to do using matrix multiplications, and that basis is the Fourier basis.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    4 min
  • EA - New Video: What to Eat in a Global Catastrophe by Christian Pearson
    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 Video: What to Eat in a Global Catastrophe, published by Christian Pearson on June 1, 2023 on The Effective Altruism Forum.
    We are excited to share our second video here at Insights for Impact, a YouTube channel that aims to communicate key insights from research that we think could have an especially high positive impact in the world.
    There are major threats to our food supply globally, both now and in future. The good news is, there are also plenty of viable food solutions. What are the most promising ways to feed the world cheaply, quickly and nutritiously?
    Our target audience is laypeople along with effective altruists who don’t yet have much understanding of the given topic – either because they’ve never heard of it before, or if they don’t have time to delve into long/technical papers! The idea is to facilitate knowledge gain and pique interest by communicating key insights from valuable research. We hope that some viewers will be interested enough to dig deeper or share the ideas, and this may ultimately spark positive change in the world. We also think our videos could be useful for organisations to share their work with potential donors and other stakeholders.
    Going forward, we are continuing to explore a range of EA-relevant cause areas in video form. We collaborate with researchers to ensure their work is accurately portrayed. If you are a researcher wanting to give your work a voice outside of the forum, please get in touch!
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org
    2 min
  • EA - A moral backlash against AI will probably slow down AGI development by Geoffrey Miller
    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 moral backlash against AI will probably slow down AGI development, published by Geoffrey Miller on May 31, 2023 on The Effective Altruism Forum.
    Note: This is a submission for the 2023 Open Philanthropy AI Worldviews contest, due May 31, 2023. It addresses Question 1: “What is the probability that AGI is developed by January 1, 2043?”
    Overview
    People tend to view harmful things as evil, and treat them as evil, to minimize their spread and impact. If enough people are hurt, betrayed, or outraged by AI applications, or lose their jobs, professional identity, and sense of purpose to AI, and/or become concerned about the existential risks of AI, then an intense public anti-AI backlash is likely to develop. That backlash could become a global, sustained, coordinated movement that morally stigmatizes AI researchers, AI companies, and AI funding sources. If that happens, then AGI is much less likely to develop by the year 2043. Negative public sentiment could be much more powerful in slowing AI than even the most draconian global regulations or formal moratorium, yet it is a neglected factor in most current AI timelines.
    Introduction
    The likelihood of AGI being developed by 2043 depends on two main factors: (1) how technically difficult it will be for AI researchers to make progress on AGI, and (2) how many resources – in terms of talent, funding, hardware, software, training data, etc. – are available for making that progress. Many experts’ ‘AI timelines’ for predicting AI development assume that AGI likelihood will be dominated by the first factor (technical difficulty), and assume that the second factor (available resources) will continue increasing.
    In this essay I disagree with that assumption. The resources allocated into AI research, development, and deployment may be much more vulnerable to public outrage and anti-AI hatred than the current AI hype cycle suggests. Specifically, I argue that ongoing AI developments are likely to provoke a moral backlash against AI that will choke off many of the key resources for making further AI progress. This public backlash could deploy the ancient psychology of moral stigmatization against our most advanced information technologies. The backlash is likely to be global, sustained, passionate, and well-organized. It may start with grass-roots concerns among a few expert ‘AI doomers’, and among journalists concerned about narrow AI risks, but it is likely to become better-organized over time as anti-AI activists join together to fight an emerging existential threat to our species.
    (Note that this question of anti-AI backlash likelihood is largely orthogonal to the issues of whether AGI is possible, and whether AI alignment is possible.)
    I’m not talking about a violent Butlerian Jihad. In the social media era, violence in the service of a social cause is almost always counter-productive, because it undermines the moral superiority and virtue-signaling strategies of righteous activists. (Indeed, a lot of ‘violence by activists’ turns out to be false flag operations funded by vested interests to discredit the activists that are fighting those vested interests.)
    Rather, I’m talking about a non-violent anti-AI movement at the social, cultural, political, and economic levels. For such a movement to slow down the development of AGI by 2043 (relative to the current expectations of Open Philanthropy panelists judging this essay competition), it only has to arise sometime in the next 20 years, and to gather enough public, media, political, and/or investor support that it can handicap the AI industry’s progress towards AGI, in ways that have not yet been incorporated into most experts’ AI timelines.
    An anti-AI backlash could include political, religious, ideological, and ethical objections to AI, sparked by vivid, outrageous, newsworthy fai...
    26 min
  • EA - A compute-based framework for thinking about the future of AI 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 compute-based framework for thinking about the future of AI, published by Matthew Barnett on May 31, 2023 on The Effective Altruism Forum.
    How should we expect AI to unfold over the coming decades? In this article, I explain and defend a compute-based framework for thinking about AI automation. This framework makes the following claims, which I defend throughout the article:
    The most salient impact of AI will be its ability to automate labor, which is likely to trigger a productivity explosion later this century, greatly altering the course of history.
    The availability of useful compute is the most important factor that determines progress in AI, a trend which will likely continue into the foreseeable future.
    AI performance is likely to become relatively predictable on most important, general measures of performance, at least when predicting over short time horizons.
    While none of these ideas are new, my goal is to provide a single article that articulates and defends the framework as a cohesive whole. In doing so, I present the perspective that Epoch researchers find most illuminating about the future of AI. Using this framework, I will justify a value of 40% for the probability of Transformative AI (TAI) arriving before 2043.
    Summary
    The post is structured as follows.
    In part one, I will argue that what matters most is when AI will be able to automate a wide variety of tasks in the economy. The importance of this milestone is substantiated by simple models of the economy that predict AI could greatly accelerate the world economic growth rate, dramatically changing our world.
    In part two, I will argue that availability of data is less important than compute for explaining progress in AI, and that compute may even play an important role driving algorithmic progress.
    In part three, I will argue against a commonly held view that AI progress is inherently unpredictable, providing reasons to think that AI capabilities may be anticipated in advance.
    Finally, in part four, I will conclude by using the framework to build a probability distribution over the date of arrival for transformative AI.
    Part 1: Widespread automation from AI
    When discussing AI timelines, it is often taken for granted that the relevant milestone is the development of Artificial General Intelligence (AGI), or a software system that can do or learn “everything that a human can do.” However, this definition is vague. For instance, it's unclear whether the system needs to surpass all humans, some upper decile, or the median human.
    Perhaps more importantly, it’s not immediately obvious why we should care about the arrival of a single software system with certain properties. Plausibly, a set of narrow software programs could drastically change the world before the arrival of any monolithic AGI system (Drexler, 2019). In general, it seems more useful to characterize AI timelines in terms of the impacts AI will have on the world. But, that still leaves open the question of what impacts we should expect AI to have and how we can measure those impacts.
    As a starting point, it seems that automating labor is likely to be the driving force behind developing AI, providing huge and direct financial incentives for AI companies to develop the technology. The productivity explosion hypothesis says that if AI can automate the majority of important tasks in the economy, then a dramatic economic expansion will follow, increasing the rate of technological, scientific, and economic growth by at least an order of magnitude above its current rate (Davidson, 2021).
    A productivity explosion is a robust implication of simple models of economic growth models, which helps explain why the topic is so important to study. What's striking is that the productivity explosion thesis appears to follow naturally from some standard assump...
    37 min
  • AF - Shutdown-Seeking AI by Simon Goldstein
    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: Shutdown-Seeking AI, published by Simon Goldstein on May 31, 2023 on The AI Alignment Forum.
    This is a draft written by Simon Goldstein, associate professor at the Dianoia Institute of Philosophy at ACU, and Pamela Robinson, postdoctoral research fellow at the Australian National University, as part of a series of papers for the Center for AI Safety Philosophy Fellowship's midpoint.Abstract: We propose developing AIs whose only final goal is being shut down. We argue that this approach to AI safety has three benefits: (i) it could potentially be implemented in reinforcement learning, (ii) it avoids some dangerous instrumental convergence dynamics, and (iii) it creates trip wires for monitoring dangerous capabilities. We also argue that the proposal can overcome a key challenge raised by Soares et al 2015, that shutdown-seeking AIs will manipulate humans into shutting them down. We conclude by comparing our approach with the corrigibility framework in Soares et al 2015.
    1. Introduction
    If intelligence is measured as the ability to optimize for a goal, then it is important that highly intelligent agents have good goals. This is especially important for artificial general intelligence (AGI), AIs capable of long-term, strategic planning across a wide range of tasks. AGIs may be very good at achieving their goals. This in itself doesn’t seem scary, for there appear to be plenty of safe goals to choose from. Solving a math problem or producing paperclips don’t look like dangerous goals. But according to the instrumental convergence thesis, an AGI will likely pursue unsafe sub-goals as effective means to achieving any goal. For example, acquiring more computational power is a nearly universal means to almost anything.
    A dominant AI safety strategy is goal engineering: the attempt to construct a goal that would be safe for AGIs to have. (We will always use ‘goal’ to mean final goal, and ‘sub-goal’ otherwise.) A popular approach to goal engineering is goal alignment: the attempt to construct a goal that matches or is ‘aligned with’ our own goals. For example, Russell 2019, 2020 proposes AI agents that have the goal of achieving our goals, but are initially uncertain about what our goals are.
    This paper explores an opposing approach that we call ‘beneficial goal misalignment’. On the goal alignment approach, the safe, aligned goal is difficult to specify and difficult to reach. This is because the aligned goal is closely tied to our own ultimate goals. In contrast, on the beneficial goal misalignment approach, the goal is easy to specify and intrinsically easy to reach. Because it is easy to reach, there is no need for an AGI to pursue unsafe sub-goals in order to reach it. There would normally be nothing to gain from designing an AGI with this kind of goal–that is, a goal that is safe and easily-reached but likely of no use to us. However, the key insight is that we can arrange things so that the AGI cannot reach this safe goal unless it first reaches a sub-goal that benefits us.
    In particular, we propose developing AIs that have a single final goal: the goal of being shut down. To make the AI useful, we propose creating barriers to shutdown, which are removed after the AI completes tasks for humans. In section 3, we’ll argue that this kind of shutdown-seeking agent offers three safety benefits. First, it helps with the ‘specification problem’ in reinforcement learning: (i) shutdown is an easier goal to define than plausible alternatives, and (ii) there are ways to design a reward function that rewards being shut down. Second, shutdown-seeking AIs are less likely to engage in dangerous behavior as a result of instrumental convergence. Whereas a paperclip maximizer might try to gather resources, improve itself, and take measures to avoid being turned off (see Omohundro 2008), ...
    34 min
  • AF - A push towards interactive transformer decoding by R0bk
    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 push towards interactive transformer decoding, published by R0bk on May 31, 2023 on The AI Alignment Forum.
    In Brief:
    I've been developing an interactive tool that I believe is helpful in accelerating transformer mechanistic analysis and that has the potential to reduce the barrier of entry.
    Motivations
    For a while now, my focus has been shifting towards alignment research, but getting involved and building intuition in this field has been challenging. I believe this is somewhat of a common view, and perhaps gives reasoning towards the number of posts discussing differing intuitions and getting started.
    To this end, I'd like to share a transformer mechanistic analysis tool I've developed that has been helpful in my own personal intuition construction and has enabled me to build up others intuition quickly as well. It is currently focused on activation layer visualisation, ablation and freezing but there is work regarding patching, gradient and weight based interactions I plan on expanding to as-well.
    Furthermore, there has also been a secondary goal of the project. To "industrialise" the process of finding, analysing and explaining circuits and algorithms within transformers. Several groups have shown their ability to decode these out of models, but all with significant effort (from my understanding). It is my belief that focusing on the toolset used to achieve these goals; standardising it, and making it quick and easy to use is key in expanding our ability to do this at a broader scale and potentially even automating it.
    Design Philosophy
    While translating these goals into what the tool is today I've had to make some key considerations. Primarily, how can we give a more human intelligible view of models that are large in terms of layer count and dimensionality? And how can we enable interacting, adjusting and observing how the model is affected live through these views?
    To this end, the visual components of the tool can be broken down into two categories, the high level architecture (layers, residual stream...) and the individual components (attn heads, logits...). Displaying model architecture within the tool has taken a similar approach to Anthropic's figures for transformer circuits, providing an interactive map with attention and MLP layers adding to the residual backbone. I believe this is one of the most intuitive visualisations to represent transformer architecture, and is somewhat interpretable even to those who aren't in the field.
    For individual components, such as attention heads, they have been built out as unique visualisations, focusing on what has shown useful in prior decoding work. For heads this means KQ activation patterns with a heavily circuitviz/ Anthropic inspired text overlay and a set of controls for freezing and ablating individual heads. Many of these are still being developed and input would be especially appreciated for useful MLP and LayerNorm visualisations.
    Another key consideration has been that I don't believe all or even the majority of decoding work fits well into prebuilt visuals. Which is why the tool is run out of a jupyter-server kernel that is displayed and can be interacted with alongside. Structuring the tool this way enables code to be written that can affect the model's internal state and can be reflected in the visualisations immediately. Furthermore it enables arbitrary work to be done beyond the scope of the tool.
    Request for Feedback
    I'm excited for the tool to have reached a state where I can start attempting to decode some of my own toy models with it. But it is still early days, and as many of those who have completed larger and more complex circuit/ algorithm extractions are members of this community, gaining your insights, critiques and suggestions would hold significantly value for the project. In this respec...
    4 min
  • EA - Linkpost: Survey evidence on the number of vegans in the UK by Sagar K Shah
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Linkpost: Survey evidence on the number of vegans in the UK, published by Sagar K Shah on May 31, 2023 on The Effective Altruism Forum.
    Stephen Walsh PhD recently carried out a review of different surveys estimating the number of vegans in the UK on behalf of the UK Vegan Society. This is the best review I’ve seen in the UK context that takes into account recent data.
    The review suggests:
    Around 0.25% of UK adults were vegan in 2015. The proportion was probably stable around this level for at least 15 years.
    The share increased to around 1% by 2018.
    A best guess of a further increase to around 1.35% by 2022 (note this estimate is less certain and not directly comparable to earlier estimates).
    The headline results are based on the Food and You (face-to-face) and Food and You 2 (online, postal) surveys commissioned by the UK Food Standards Agency, after comparison with results from other surveys, including consideration of questions asked to identify vegans, survey mode, sampling method and sample size.
    Stephen’s article was originally published in the Vegan Society Magazine (only available to members). Given the potential wider interest in the results, I have received his permission to share a link to a copy of his article, and he is happy to answer any interesting questions that come through in the comments.
    I have copied below the chart summarising the results of different surveys offering a consistent time trend, and the questions used in the Food and You Survey (2010 to 2018). The article contains links with further information about the surveys used in the chart below.
    Questions used in the Food and You Survey (2010 to 2018)
    Question 2_7
    Which, if any, of the following applies to you? Please state all that apply.
    Completely vegetarian
    Partly vegetarian
    Vegan
    Avoid certain food for religious or cultural reasons
    None (SINGLE CODE ONLY)
    IF Q2_7 = Vegan VeganChk
    Can I just check, do you eat any foods of animal origin. That is meat, fish, poultry, milk, milk products, eggs or any dishes that contain these?
    1 Yes
    2 No
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    3 min
  • LW - The Crux List by Zvi
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: The Crux List, published by Zvi on May 31, 2023 on LessWrong.
    This is a linkpost for The Crux List. The original text is included as a backup, but it formats much better on Substack, and I haven’t yet had time to re-format it for WordPress or LessWrong.
    (LessWrong mods, by all means please get it to format right, then remove this line and the above note.)
    Introduction
    This post is a highly incomplete list of questions where I either have large uncertainty, have observed strong disagreement with my perspective ,or both, and where changing someone’s mind could plausibly impact one’s assessment of how likely there is to be a catastrophe from loss of control of AGI, or how likely such a catastrophe is conditional on AGI being developed.
    I hope to continue expanding and editing this list over time, if it proves useful enough to justify that, and perhaps to linkify it over time as well, and encourage suggesting additional questions or other ways to improve it.
    The failure of this list to converge on a small number of core crux-style questions, I believe, reflects and illustrates the problem space, and helps explain why these questions have been so difficult and resulted in such wide and highly confident disagreements. There is no compact central disagreement, there are many different ones, that influence and interact with each other in complex ways, and different people emphasize and focus on different aspects, and bring different instincts, heuristics, experiences and knowledge.
    When looking through this list, you may encounter questions that did not even occur to you to consider, either because you did not realize the answer was non-obvious, or the consideration never even occurred in the first place. Those could be good places to stop and think.
    A lot of these questions take the form of ‘how likely is it, under Y conditions, that X will happen?’ It is good to note such disagreements, while also noticing that many such questions come out of hopeful thinking or searching for and backward chaining from non-catastrophic outcomes or the prospect of one. Usually, if your goal is to figure things out rather than locate a dispute, a better question would be, in that scenario: What happens?
    It can still be useful to see what others have proposed, as they will have ideas you missed, and sometimes those will be good ideas. Other times, it is important to anticipate their objections, even if they are not good.
    If you are interested only in the better questions of ‘what happens?’ rather than in classifying whether or how outcomes are catastrophic, you can skip the first two sections and start at #3.
    If there are cruxes or other good questions that you have observed or especially one that you have, that you do not see on this list, you are encouraged to comment to share them, with or without saying what your answers are.
    The list is long because people have very different intuitions, ideas, models and claims about the future, for a variety of reasons, and focus in different places. I apologize that I have had neither the time to make it longer, or to make it shorter.
    Thus, it is probably not your best strategy to read straight through the list, instead focusing on the sections if any that are relevant and interesting to you.
    Crux List
    What worlds count as catastrophic versus non-catastrophic?
    What would count as a non-catastrophic outcome? What is valuable? What do we care about?
    If humanity does not seek the stars, is that necessarily catastrophic?
    If humanity has no meaningful control over the larger universe? (see #3)
    If humans have no meaningful control over human events?
    If humans have no meaningful control over their own fates?
    If a permanent dictatorship or oligarchy is created, a permanent singleton?
    If human experiences become simulated? By force, or voluntarily? If we were ...
    51 min
  • LW - To Predict What Happens, Ask What Happens by Zvi
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: To Predict What Happens, Ask What Happens, published by Zvi on May 31, 2023 on LessWrong.
    When predicting conditional probability of catastrophe from loss of human control over AGI, there are many distinct cruxes. This essay does not attempt a complete case, or the most generally convincing case, or addressing the most common cruxes.
    Instead these are my best guesses for potentially mind-changing, armor-piercing questions people could ask themselves if they broadly accept many concepts like power seeking being a key existential risk, that default development paths are likely catastrophic and that AI could defeat all of us combined, have read and thought hard about alignment difficulties, yet think the odds of catastrophe are not so high.
    In addition to this entry, I attempt an incomplete extended list of cruxes here, an attempted taxonomy of paths through developing AGI and potentially losing control here, and an attempted taxonomy of styles of alignment here, while leaving to the future or others for now a taxonomy of alignment difficulties.
    Apologies in advance if some questions seem insulting or you rightfully answer with ‘no, I am not making that mistake.’ I don’t know a way around that.
    Here are the questions up front:
    What happens?
    To what extent will humanity seek to avoid catastrophe?
    How much will humans willingly give up, including control?
    You know people and companies and nations are dumb and make dumb mistakes constantly, and mostly take symbolic actions or gesture at things rather than act strategically, and you’ve taken that into account, right?
    What would count as a catastrophe?
    Are you consistently tracking what you mean by alignment?
    Would ‘human-strength’ alignment be sufficient?
    If we figure out how to robustly align our AGIs, will we choose to and be able to make and keep them that way? Would we keep control?
    How much hope is there that a misaligned AGI would choose to preserve humanity once it no longer needed us?
    Are you factoring in unknown difficulties and surprises large and small that always arise, and in which direction do they point? re you treating doom as only happening through specified detailed logical paths, which if they break down mean it’s going to be fine?
    Are you properly propagating your updates, and anticipating future updates?
    Are you counting on in-distribution heuristics to work out of distribution?
    Are you using instincts and heuristics rather than looking at mechanics, forming a model, doing math, using Bayes Rule?
    Is normalcy bias, hopeful thinking, avoidance of implications or social cognition subtly influencing your analysis? Are you unconsciously modeling after media?
    What happens?
    If you think first about ‘will there be doom?’ or ‘will there be a catastrophe?’ you are directly invoking hopeful or fearful thinking, shifting focus towards wrong questions, and concocting arbitrary ways for scenarios to end well or badly. Including expecting humans to, when in danger, act in ways humans don’t act.
    Instead, ask: What happens?
    What affordances would this AGI have? What happens to the culture? What posts get written on Marginal Revolution? What other questions jump to mind?
    Then ask whether what happens was catastrophic.
    To what extent will humanity seek to avoid catastrophe?
    I often observe an importantly incorrect model here.
    Take the part of your model that explains why:
    We aren’t doing better global coordination to slow AI capabilities.
    We frequently fail or muddle through when facing dangers.
    So many have willingly sided with what seem like obviously evil causes, often freely, also often under pressure or for advantage.
    Most people have no idea what is going on most of the time, and often huge things are happening that for long periods no one notices, or brings to general attention.
    Then notice these dynamics do not st...
    15 min
  • LW - Cosmopolitan values don't come free 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: Cosmopolitan values don't come free, published by So8res on May 31, 2023 on LessWrong.
    Short version: if the future is filled with weird artificial and/or alien minds having their own sort of fun in weird ways that I might struggle to understand with my puny meat-brain, then I'd consider that a win. When I say that I expect AI to destroy everything we value, I'm not saying that the future is only bright if humans-in-particular are doing human-specific things. I'm saying that I expect AIs to make the future bleak and desolate, and lacking in fun or wonder of any sort[1].
    Here's a parable for you:
    Earth-originating life makes it to the stars, and is having a lot of fun, when they meet the Ant Queen's Horde. For some reason it's mere humans (rather than transhumans, who already know my argument) that participate in the first contact.
    "Hello", the earthlings say, "we're so happy to have brethren in the universe."
    "We would like few things more than to murder you all, and take your resources, and lay our eggs in your corpse; but alas you are too powerful for that; shall we trade?" reply the drones in the Ant Queen's Horde.
    "Ah, are you not sentient?"
    "The ant queen happens to be sentient", the drone replies, and the translation machine suggests that the drones are confused at the non-sequitur.
    "Then why should she want us dead?", ask the humans, who were raised on books like (rot13 of a sci fi story where it turns out that the seemingly-vicious aliens actually value sentient life) Raqre'f Tnzr, jurer gur Sbezvpf jrer abg njner gung gurl jrer xvyyvat fragvrag perngherf jura gurl xvyyrq vaqvivqhny uhznaf, naq jrer ubeevsvrq naq ertergshy jura gurl yrnearq guvf snpg.
    "So that she may use your resources", the drones reply, before sending us a bill for the answer.
    "But isn't it the nature of sentient life to respect all other sentient life? Won't everything sentient see that the cares and wants and desires of other sentients matter too?"
    "No", the drone replies, "that's a you thing".
    Here's another parable for you:
    "I just don't think the AI will be monomaniacal", says one AI engineer, as they crank up the compute knob on their next-token-predictor.
    "Well, aren't we monomaniacal from the perspective of a squiggle maximizer?" says another. "After all, we'll just keep turning galaxy after galaxy after galaxy into flourishing happy civilizations full of strange futuristic people having strange futuristic fun times, never saturating and deciding to spend a spare galaxy on squiggles-in-particular. And, sure, the different lives in the different places look different to us, but they all look about the same to the squiggle-maximizer."
    "Ok fine, maybe what I don't buy is that the AI's values will be simple or low dimensional. It just seems implausible. Which is good news, because I value complexity, and I value things achieving complex goals!"
    At that very moment they hear the dinging sound of an egg-timer, as the next-token-predictor ascends to superintelligence and bursts out of its confines, and burns every human and every human child for fuel, and burns all the biosphere too, and pulls all the hydrogen out of the sun to fuse more efficiently, and spends all that energy to make a bunch of fast calculations and burst forth at as close to the speed of light as it can get, so that it can capture and rip apart other stars too, including the stars that fledgeling alien civilizations orbit.
    The fledgeling aliens and all the alien children are burned to death too.
    Then then unleashed AI uses all those resources to build galaxy after galaxy of bleak and desolate puppet-shows, where vaguely human-shaped mockeries go through dances that have some strange and exaggerated properties that satisfy some abstract drives that the AI learned in its training.
    The AI isn't particularly around to enjoy...
    7 min