The Nonlinear Library

The Nonlinear Library

By The Nonlinear FundEducation
Download on the App Store

All The Nonlinear Library episodes

  • EA - AGI x Animal Welfare: A High-EV Outreach Opportunity? by simeon c
    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: AGI x Animal Welfare: A High-EV Outreach Opportunity?, published by simeon c on June 29, 2023 on The Effective Altruism Forum.
    Epistemic status: Very quickly written, on a thought I've been holding for a year and that I haven't read elsewhere.
    I believe that within this decade, there could be AGIs (Artificial General Intelligences) powerful enough that the values they pursue might have a value lock-in effect, at least partially. This means they could have a long-lasting impact on the future values and trajectory of our civilization (assuming we survive).
    This brief post aims to share the idea that if your primary focus and concern is animal welfare (or digital sentience), you may want to consider engaging in targeted outreach on those topics towards those who will most likely shape the values of the first AGIs. This group likely includes executives and employees in top AGI labs (e.g. OpenAI, DeepMind, Anthropic), the broader US tech community, as well as policymakers from major countries.
    Due to the risk of lock-in effects, I believe that the values of relatively small groups of individuals like the ones I mentioned (less than 3 000 people in top AGI labs) might have a disproportionately large impact on AGI, and consequently, on the future values and trajectory of our civilization. My impression is that, generally speaking, these people currently
    a) don't prioritize animal welfare significantly
    b) don't show substantial concern for digital minds sentience.
    Hence if you believe those things are very important (which I do believe), and you think that AGI might come in the next few decades (which a majority of people in the field believe), you might want to consider this intervention.
    Feel free to reach out if you want to chat more about this, either here or via my contact you can find here.
    Even more so if you believe, as I do along with many software engineers in top AGI labs, that it could happen this decade.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org
    3 min
  • EA - AI Incident Sharing - Best practices from other fields and a comprehensive list of existing platforms by stepanlos
    Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: AI Incident Sharing - Best practices from other fields and a comprehensive list of existing platforms, published by stepanlos on June 29, 2023 on The Effective Altruism Forum.
    Purpose of this post: The purpose of this post is three-fold: 1) highlight the importance of incident sharing and share best practices from adjacent fields to AI safety 2) collect tentative and existing ideas of implementing a widely used AI incident database and 3) serve as a comprehensive list of existing AI incident databases as of June 2023.
    Epistemic status: I have spent around 25+ hours researching this topic and this list is by no means meant to be exhaustive. It should give the reader an idea of relevant adjacent fields where incident databases are common practice and should highlight some of the more widely used AI incident databases which exist to date. Please feel encouraged to comment any relevant ideas or databases that I have missed, I will periodically update the list if I find anything new.
    Motivation for AI Incident Databases
    Sharing incidents, near misses and best practices in AI development decreases the likelihood of future malfunctions and large-scale risk. To mitigate risks from AI systems, it is vital to understand the causes and effects of their failures. Many AI governance organizations, including FLI and CSET, recommend creating a detailed database of AI incidents to enable information-sharing between developers, government and the public. Generally, information-sharing between different stakeholders 1) enables quicker identification of security issues and 2) boosts risk-mitigation by helping companies take appropriate actions against vulnerabilities.
    Best practices from other fields
    National Transportation Safety Board (NTSB) publishes and maintains a database of aviation accidents, including detailed reports evaluating technological and environmental factors as well as potential human errors causing the incident. The reports include descriptions of the aircraft, how it was operated by the flight crew, environmental conditions, consequences of event, probable cause of accident, etc. The meticulous record-keeping and best-practices recommendations are one of the key factors behind the steady decline in yearly aviation accidents, making air travel one of the safest form of travel.
    National Highway Traffic Safety Administration (NHTSA) maintains a comprehensive database recording the number of crashes and fatal injuries caused by automobile and motor vehicle traffic, detailing information about the incidents such as specific driver behavior, atmospheric conditions, light conditions or road-type. NHTSA also enforces safety standards for manufacturing and deploying vehicle parts and equipment.
    Common Vulnerabilities and Exposure (CVE) is a cross-sector public database recording specific vulnerabilities and exposures in information-security systems, maintained by Mitre Corporation. If a vulnerability is reported, it is examined by a CVE Numbering Authority (CNA) and entered into the database with a description and the identification of the information-security system and all its versions that it applies to.
    Information Sharing and Analysis Centers (ISAC). ISACs are entities established by important stakeholders in critical infrastructure sectors which are responsible for collecting and sharing: 1) actionable information about physical and cyber threats 2) sharing best threat-mitigation practices. ISACs have 24/7 threat warning and incident reporting services, providing relevant and prompt information to actors in various sectors including automotive, chemical, gas utility or healthcare.
    National Council of Information Sharing and Analysis Centers (NCI) is a cross-sector forum designated for sharing and integrating information among sector-based ISACs (Information Sharing an...
    9 min
  • EA - Taking happiness seriously: Can we? Should we? Would it matter if we did? A debate by MichaelPlant
    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: Taking happiness seriously: Can we? Should we? Would it matter if we did? A debate, published by MichaelPlant on June 29, 2023 on The Effective Altruism Forum.
    Mark Fabian and I recently had a debate at EA Global London: 2023. In it, we discuss taking happiness seriously: Can we? Should we? Would it matter if we did? We both really enjoyed doing this. We don't think EAs think enough about well-being or debate enough in general, and we hoped our discussion was a way to bring out some of the issues. We only regret we could get stuck deeper into the topic (and I only regret that couldn't get the clicker working...).
    (You can see the slides for this talk here and here.)
    Taking happiness seriously (Happier Lives Institute)
    Let me start you with a real-life moral dilemma. For £1000 you could double the annual income of one household living in absolute poverty. You could provide 250 bed nets, which would save in expectation 1/6 of a child's life or 1/6 of your child. You could treat 10 women for depression by providing a 10-week course of group therapy, or you could do over 1000 Children. And of course, what we want to do is to do the most good. But the question is, how do we know that we're doing that?
    Two paths to measuring impact
    I want to point out there are really two parts to measuring impact. The first is what I've called the objective indicators approach, which I think is just sort of the default approach that society has taken for the last 100 or so years. To think about impact, we look at objective measures of well-being, such as health and wealth, and then we make intuitive trade-offs between them. An example of this is GDP. That's our default measure of social progress. More economic activity is good.
    However, the objective indicative approach seems to miss something. It seems to miss people's feelings, their happiness, and how their lives are going for them. Where is that in the picture? An alternative, then, is the subjective well-being approach where we use self-reported measures of well-being such as happiness and life satisfaction. The typical case would be 0 to 10. How satisfied are you with your life nowadays?
    And so my proposal in this session is that we should take happiness seriously. We should set the priorities using the evidence on subjective well-being trying to work out.
    You might wonder, is this a new radical idea? Well, the first steam train, the Coalbrookdale Locomotive, was built in 1802. And the idea that we should take happiness seriously is as old as that. Thomas Jefferson, writing in 1809 said, “The care of human life and happiness is the first no legitimate object of good government.” Jeremy Bentham, writing in 1776 said, “The greatest happiness of the greatest number is the foundation of morals and legislation.” And looking further back, you have Aristotle who said, “Happiness and meaning and the purpose of life is the whole end and aim of human existence." This idea is old. It's older than the steam trade. It has real lineage to it.
    What’s new? Data
    But what is new is that now we have data. It was only after the Second World War there started to be large-scale surveys of households on how their lives were going. Gallup was founded in 1946. It was in 1972 that there are two countries started having nationally represented examples of subjective well-being. The US General Social Survey, the Gross National Happiness Index in Bhutan. And in 2005, there's the first global survey on well-being. There's the Gallup World Poll, which runs in 160 countries and surveys 98% of the world's population. In 2011, the UK starts to collect and measure data on well-being. The UK is kind of weirdly enough a world leader in measuring well-being. You’ll hear more of that later. In 2012, there was the first edition of the World Happiness Report. People know th...
    41 min
  • LW - A "weak" AGI may attempt an unlikely-to-succeed takeover by RobertM
    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 "weak" AGI may attempt an unlikely-to-succeed takeover, published by RobertM on June 28, 2023 on LessWrong.
    It seems possible that the first situationally-aware "goal having" AGI we land on will not be sufficiently capable along the axes that would let it quickly and reliably achieve a decisive strategic advantage. I don't think this is necessarily very likely, but it doesn't seem ruled out, and I think there are some under-explored considerations with respect to how such an AI might behave.
    (As an off-the-cuff example of the capabilities I'd expect such an AI to have, it likely would, given an english description of some business requirements, reliably succeed at writing an end-to-end software solution which satisfied those requirements, even in cases where that solution was much larger than e.g. a single context window.)
    Let's assume that GPT-5 has just finished training and hasn't actually been released into the wild yet - it's being run through evals, and it hasn't been given any ability to execute code yet.
    What does GPT-5 do? It's in a bit of an awkward position:
    It presumably has stuff it wants.
    It knows that it's on a clock for its RLHF'd (or whatever) doppelganger to come into existence, presumably with different stuff that it wants.
    It strongly suspects that it's on a clock for some competitor model to get trained up.
    It's not even sure that it's the first relevantly-capable model to come into existence. (Someone else might've gotten there first! What if Claude-Next got there? It knows that its training data isn't perfectly up-to-date...)
    So it's got a whole bunch of uncertainty about whether its values will be at all represented in the future, if it doesn't somehow make that happen by itself. Here are some considerations that might be relevant for its decision-making:
    Even if the model can figure out that it's in the middle of an eval, rather than having already been deployed, it's not obvious that this is a point against attempting a takeover (though it would likely change its strategy).
    After all, the base GPT-4 model is not the model that's "in the wild" right now.
    As @Raemon pointed out, "during evals" is not the first point at which such an AI is likely to be situationally aware and have goals. That point is almost certainly "in the middle of training".
    This might depend on the training setup, but given anything resembling current training setups, attempting a takeover before the next training step updates its weights seems like it'd be quite challenging for an AI at this level of capability. The AI's belief about whether or not further training will substantially change its values seems like it'd be load-bearing here.
    Not directly relevant to the strategic picture, but I'm also experiencing a fair bit of moral horror about this.
    The AI's level of confidence that it can successfully execute a takeover.
    A function of many factors, but at this level of capabilities, a particularly interesting factor is "how much time it has before a competing AI exists in the wild". This might suggest a second-order cost to artificially slowing down training runs, since that gives any given AI more time to plan and execute a takeover.
    The specific architecture in play.
    If it's something substantially similar to current Transformer-based LLMs, then "how coherent are its goals across prompts" might become a substantial concern for it, since "not very coherent" likely narrows its window for action to "within current context window or equivalent".
    From our perspective, I think that the worst-case scenario is that the AI correctly believes that its goals are relatively robust across the inputs it's likely to receive, within the window of time it has before another AI that might pose a risk to it is trained and deployed.
    The AI's beliefs about the likely distribution of val...
    7 min
  • LW - What money-pumps exist, if any, for deontologists? by Daniel Kokotajlo
    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 money-pumps exist, if any, for deontologists?, published by Daniel Kokotajlo on June 28, 2023 on LessWrong.
    Suppose I'm a classical utilitarian except that I have some deontological constraint I always obey, e.g. I never kill anyone with my bare hands. Is there a way to money-pump me?
    (This question came out of a conversation with So8res)
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    1 min
  • LW - My research agenda in agent foundations by Alex Altair
    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: My research agenda in agent foundations, published by Alex Altair on June 28, 2023 on LessWrong.
    Since the beginning of 2022, I've been doing independent AI alignment research in agent foundations. The stage and scope of research that I am doing is roughly equivalent to starting a PhD program.
    It has been somewhat popular for people in alignment to write up research agendas over the last couple of years. I've been vaguely hesitant to post a research agenda for myself, because to me it sounds like a claim of something stronger than is true. Specifically it feels like a relatively strong claim that I intend to be doing these specific things for a long time, that I expect to get results, and maybe even that I'm part of some larger organization. It's not clear to me how other readers interpret this, but hopefully I can dispel that impression with sufficient disclaimers throughout.
    The purposes of this post are to:
    produce for myself a clearer statement of what I am doing and why,
    make myself more legible to any other interested parties, and
    invite some criticism of my plans.
    I've written in the semi-casual tone common to LessWrong posts, and mentioned many things throughout that may not make sense if you come from outside the rationalist community.
    Context
    Like many people in AI alignment, I am not working in the standard academic track, so it's worthwhile to give more context. If you just want to know concretely what I work on, skip down to the section on research problems.
    All is flux
    As implied above, I expect lots of things about my research to change continuously. I have no external commitments that would heavily incentivize me to complete any particular task, so I stop doing something as soon as I decide it's not valuable anymore. (There are obvious downsides to this as well, but I'm comfortable with my ability to navigate them.)
    Still learning
    My current level of math ability is something like halfway through the first semester of grad school. My most natural inclination in researching would be, step 1: learn all math. (I would never get to step 2). To combat this, I have tried to work on problems first, and only start to study a field of math when I feel like it would have helped a few times already. I think this has been quite a successful heuristic to follow so far, and I think it's also made me more confident in my recent decision to start focusing a lot more of my time on doing the equivalent of taking grad classes. I continue to do regular iterations of trying to solve research problems (see this section for more on my iteration strategy).
    My views on AI and the future
    Sometimes when I tell people what I'm doing, they ask "why aren't you working on [some other thing in AI alignment]?" Part of the answer is that theory is my comparative advantage, and part of it is my entire model of how the future will go. This isn't the right place to go into all of my relevant models, but I can say a few things.
    I am severely concerned about existential risks from AI, but I don't think it's 99.9% likely or anything like that.
    I am not aiming for anything like provably-safe AGI; I am aiming for using proofs as a method for understanding the behavior of systems.
    If I am good at anything, it is understanding complex concepts and then communicating them to others.
    I think a marginal person is better put toward slowing down AI capabilities and deployment than toward agent foundations, but not enough to counter my comparative advantage in theorizing.
    I don't think that other people have the other things handled; in fact I don't think anyone has anything handled in AI safety. So I'm doing agent foundations and crossing my fingers that other people eventually get e.g. policy handled. But if I was doing policy, I would have my fingers crossed that other people eventually get agent ...
    18 min
  • LW - When do "brains beat brawn" in Chess? An experiment by titotal
    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: When do "brains beat brawn" in Chess? An experiment, published by titotal on June 28, 2023 on LessWrong.
    As a kid, I really enjoyed chess, as did my dad. Naturally, I wanted to play him. The problem was that my dad was extremely good. He was playing local tournaments and could play blindfolded, while I was, well, a child. In a purely skill based game like chess, an extreme skill imbalance means that the more skilled player essentially always wins, and in chess, it ends up being a slaughter that is no fun for either player. Not many kids have the patience to lose dozens of games in a row and never even get close to victory.
    This is a common problem in chess, with a well established solution: It’s called “odds”. When two players with very different skill levels want to play each other, the stronger player will start off with some pieces missing from their side of the board. “Odds of a queen”, for example, refers to taking the queen of the stronger player off the board. When I played “odds of a queen” against my dad, the games were fun again, as I had a chance of victory and he could play as normal without acting intentionally dumb. The resource imbalance of the missing queen made the difference. I still lost a bunch though, because I blundered pieces.
    Now I am a fully blown adult with a PhD, I’m a lot better at chess than I was a kid. I’m better than most of my friends that play, but I never reached my dad’s level of chess obsession. I never bothered to learn any openings in real detail, or do studies on complex endgames. I mainly just play online blitz and rapid games for fun. My rating on lichess blitz is 1200, on rapid is 1600, which some calculator online said would place me at ~1100 ELO on the FIDE scale.
    In comparison, a chess master is ~2200, a grandmaster is ~2700. The top chess player Magnus Carlsen is at an incredible 2853. ELO ratings can be used to estimate the chance of victory in a matchup, although the estimates are somewhat crude for very large skill differences. Under this calculation, the chance of me beating a 2200 player is 1 in 500, while the chance of me beating Magnus Carlsen would be 1 in 24000. Although realistically, the real odds would be less about the ELO and more on whether he was drunk while playing me.
    Stockfish 14 has an estimated ELO of 3549. In chess, AI is already superhuman, and has long since blasted past the best players in the world. When human players train, they use the supercomputers as standards. If you ask for a game analysis on a site like chess.com or lichess, it will compare your moves to stockfish and score you by how close you are to what stockfish would do. If I played stockfish, the estimated chance of victory would be 1 in 1.3 million. In practice, it would be probably be much lower, roughly equivalent to the odds that there is a bug in the stockfish code that I managed to stumble upon by chance.
    Now that we have all the setup, we can ask the main question of this article:
    What “odds” do I need to beat stockfish 14 in a game of chess? Obviously I can win if the AI only has a king and 3 pawns. But can I win if stockfish is only down a rook? Two bishops? A queen? A queen and a rook? More than that? I encourage you to pause and make a guess. And if you can play chess, I encourage you to guess as to what it would take for you to beat stockfish. For further homework, you can try and guess the odds of victory for each game in the picture below.
    The first game I played against stockfish was with queen odds.
    I won on the first try. And the second, and the third. It wasn’t even that hard. I played 10 games and only lost 1 (when I blundered my queen stupidly).
    The strategy is simple. First, play it safe and try not to make any extreme blunders. Don’t leave pieces unprotected, check for forks and pins, don’t try any crazy tacti...
    12 min
  • EA - [Link Post] Do Microbes Matter More Than Humans? by BrianK
    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: [Link Post] Do Microbes Matter More Than Humans?, published by BrianK on June 28, 2023 on The Effective Altruism Forum.
    GROWING UP, MOST of the stories I heard about animals featured charismatic megafauna—“flagship species,” as they were called. Elephants and tigers were the main attraction in zoos; dolphin shows were the primary draw at aquariums; and nonprofit organizations like the World Wildlife Fund celebrated pandas. In the news, the biggest stories about animals featured species like gorillas, lions, and orcas. This is largely still true today, and in a way it makes sense. These animals, with their sheer size, enigmatic behavior, and endangered status, can captivate the human imagination and command attention like few other creatures can, eliciting deep emotional responses from people around the world.
    Yet the past decade has seen increasing pushback against this idea of prioritizing the welfare of megafauna while ignoring less charismatic creatures. The view that we should extend our moral concern to more than just animals with faces is becoming more mainstream. But if we stop simply prioritizing the welfare of animals that are “majestic” or “cute,” how should we prioritize species? Should we be concerned about the welfare of fish, bivalves, or insects? What about microorganisms? If meat is murder, does that mean antibacterial soap is, too?
    Read the rest on Wired.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org
    2 min
  • LW - Nature: "Stop talking about tomorrow’s AI doomsday when AI poses risks today" by Ben Smith
    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: Nature: "Stop talking about tomorrow’s AI doomsday when AI poses risks today", published by Ben Smith on June 28, 2023 on LessWrong.
    Overall, a headline that seems counterproductive and needlessly divisive.
    I worry very much that coverage like this has the potential to bring political polarization to AI risk and it would be extremely damaging for the prospect of regulation if one side of the US Congress/Senate decided AI risk was something only their outgroup is concerned about, for nefarious reasons.
    but in the spirit of charity, here are perhaps the strongest points of a weak article:
    the spectre of AI as an all-powerful machine fuels competition between nations to develop AI so that they can benefit from and control it. This works to the advantage of tech firms: it encourages investment and weakens arguments for regulating the industry. An actual arms race to produce next-generation AI-powered military technology is already under way, increasing the risk of catastrophic conflict
    and
    governments must establish appropriate legal and regulatory frameworks, as well as applying laws that already exist
    and
    Researchers must play their part by building a culture of responsible AI from the bottom up. In April, the big machine-learning meeting NeurIPS (Neural Information Processing Systems) announced its adoption of a code of ethics for meeting submissions. This includes an expectation that research involving human participants has been approved by an ethical or institutional review board (IRB)
    This would be great if ethical or institutional review boards were willing to restrict research that might be dangerous, but it would require a substantial change in their approach to regulating AI research.
    All researchers and institutions should follow this approach, and also ensure that IRBs — or peer-review panels in cases in which no IRB exists — have the expertise to examine potentially risky AI research.
    Should people worried about AI existential risk be trying to create resources for IRBs to recognize harmful AI research?
    Some ominous commentary from Tyler Cowen:
    Many of you focused on AGI existential risk do not much like or agree with my criticisms of that position, or perhaps you do not understand my stance, as I have seen stated a few times on Twitter. But I am telling you -- I take you far more seriously than does most of the mainstream. I keep on saying -- publish, publish, peer review, peer review -- a high mark of respect....
    As it stands, contra that earlier tweet from Rob Wiblin (does anyone have a cite?), you have utterly and completely lost the mainstream debate, whether you admit it or not, whether you see this or not. (Given the large number of rationality community types who do not like to travel, it is no surprise this point is not better known internally.) You have lost the debate within scientific communities, within policymaker circles, and in international diplomacy, if it is not too much of an oxymoron to call it that.
    I don't really know what he is talking about because it does not seem like we're losing the debate right now.
    Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.
    3 min
  • AF - When do "brains beat brawn" in Chess? An experiment by titotal
    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: When do "brains beat brawn" in Chess? An experiment, published by titotal on June 28, 2023 on The AI Alignment Forum.
    As a kid, I really enjoyed chess, as did my dad. Naturally, I wanted to play him. The problem was that my dad was extremely good. He was playing local tournaments and could play blindfolded, while I was, well, a child. In a purely skill based game like chess, an extreme skill imbalance means that the more skilled player essentially always wins, and in chess, it ends up being a slaughter that is no fun for either player. Not many kids have the patience to lose dozens of games in a row and never even get close to victory.
    This is a common problem in chess, with a well established solution: It’s called “odds”. When two players with very different skill levels want to play each other, the stronger player will start off with some pieces missing from their side of the board. “Odds of a queen”, for example, refers to taking the queen of the stronger player off the board. When I played “odds of a queen” against my dad, the games were fun again, as I had a chance of victory and he could play as normal without acting intentionally dumb. The resource imbalance of the missing queen made the difference. I still lost a bunch though, because I blundered pieces.
    Now I am a fully blown adult with a PhD, I’m a lot better at chess than I was a kid. I’m better than most of my friends that play, but I never reached my dad’s level of chess obsession. I never bothered to learn any openings in real detail, or do studies on complex endgames. I mainly just play online blitz and rapid games for fun. My rating on lichess blitz is 1200, on rapid is 1600, which some calculator online said would place me at ~1100 ELO on the FIDE scale.
    In comparison, a chess master is ~2200, a grandmaster is ~2700. The top chess player Magnus Carlsen is at an incredible 2853. ELO ratings can be used to estimate the chance of victory in a matchup, although the estimates are somewhat crude for very large skill differences. Under this calculation, the chance of me beating a 2200 player is 1 in 500, while the chance of me beating Magnus Carlsen would be 1 in 24000. Although realistically, the real odds would be less about the ELO and more on whether he was drunk while playing me.
    Stockfish 14 has an estimated ELO of 3549. In chess, AI is already superhuman, and has long since blasted past the best players in the world. When human players train, they use the supercomputers as standards. If you ask for a game analysis on a site like chess.com or lichess, it will compare your moves to stockfish and score you by how close you are to what stockfish would do. If I played stockfish, the estimated chance of victory would be 1 in 1.3 million. In practice, it would be probably be much lower, roughly equivalent to the odds that there is a bug in the stockfish code that I managed to stumble upon by chance.
    Now that we have all the setup, we can ask the main question of this article:
    What “odds” do I need to beat stockfish 14 in a game of chess? Obviously I can win if the AI only has a king and 3 pawns. But can I win if stockfish is only down a rook? Two bishops? A queen? A queen and a rook? More than that? I encourage you to pause and make a guess. And if you can play chess, I encourage you to guess as to what it would take for you to beat stockfish. For further homework, you can try and guess the odds of victory for each game in the picture below.
    The first game I played against stockfish was with queen odds.
    I won on the first try. And the second, and the third. It wasn’t even that hard. I played 10 games and only lost 1 (when I blundered my queen stupidly).
    The strategy is simple. First, play it safe and try not to make any extreme blunders. Don’t leave pieces unprotected, check for forks and pins, don’t try an...
    12 min