2024 #AcademicRunPlaylist Wrapped
It’s been a great year for my #AcademicRunPlaylist! This year in addition to listening to over 2000 talks while breaking my daily running record twice, hitting 46.5 miles (75 km) in Japan in April and 49.2 miles (~79 km) in Madrid in September. I’m so grateful that I get to explore interesting places while getting some exercise and getting exposed to fascinating ideas. After exhausting talks on the channels I’ve subscribed to over the summer I also started listening to books on my runs, so here I’ll also pick my top books for the year.
As an aside, I still prefer talks to books. While there are some books whose contents aren’t well covered in academic talks (looking at you corporate law), there’s a ton of academic research that is very poorly represented in books. This is especially true of technical topics. I’m still glad to mix books in, but they’re always lower priority than new talks.
As with last year, there’s no hard methodology here. Most of these talks are ones that I still remember well months later, and the books I chose are the cream of the crop (although most weren’t published this year). Every talk here is one that I can strongly recommend, and if you’re even remotely interested in these areas they’re worth your time. The order here is mostly reverse chronological, and I’ve included abstracts and/or short commentary about why these talks are so good. The categories below are somewhat arbitrary - many times talks could easily be plugged into multiple buckets.
For those who are interested, I've updated my playlist development methodology post to include how I choose books. My entire YouTube playlist from the last ~3 years is also here (I think playlists have a limit of 5000, so I'll soon have to shift to a new playlist). Finally, for comparison's sake you might want to check out last year's wrapped post.
On to the list!
Technically Curious - AI, HCI, Robots, etc.
Shannon Vallor - Can We Live with AI?
This is a great talk on what current AI is (a mirror), what it isn't (what its boosters say it is), and what it could be. Vallor critically examines the current class of AI technologies and its capabilities, exposing it for what it truly is and the psychological tricks the sector plays to hype up its efficacy.
Callum Cant, Kirsten Sehnbruch, and James Muldoon - The Human Labor Underlying AI
This panel digs deep into the exploitative labor system underlying pretty much every large AI system out there, and I can't wait to read their book on this.
Sara Beery - Distribution shift in ecological Data
In this can’t miss talk Beery dissects general problems with distribution shift research and the illusion of recent progress in the space.
Victor Veitch - On Off-Target Behavior in LLM Alignment
Veitch provides a nice formalism for alignment, showing how one can unify different aspects of alignment to train more desirable models.
Jacob Steinhardt - Scalably Understanding AI
Steinhardt presents an impressive method to find prompt inputs that give desired model outputs, demonstrating that this method can transfer from open to closed models.
Jas Brooks - Augmented Breathing via Thermal Feedback in the Nose
I originally scoffed at the title of this talk, but I’m glad I watched it. Jas shows how by manipulating in-nose temperature one can change breath perception, enabling a fascinating variety of new interactions.
Daniel Ho - Assessing the Reliability of AI in Legal Research
Ho systematically evaluates both consumer and law-specific LLMs and finds ridiculously high error rates, even with the most "advanced" models. Hopefully this pours more cold water on people using LLMs in areas where the truth matters.
Marc Miskin - Microscopic Robots that Sense, Think, Act, and Compute
Fabricating millions of unique microscopic robots using standard equipment? Microrobots that can freaking GROW METAL APPENDAGES when they want? Microbots that WALK ON AXONS AND PULL SEVERED NERVES TOGETHER?!?!? Unbelievably sci-fi stuff, some of the coolest things happening in robotics right now IMO.
Evan Shieh and Faye-Marie Vassel - Intersectional Biases in Generative Language Models and Their Psychosocial Impacts
The herculean effort this team undertook to methodically document the nearly overwhelming biases of these models is a call to action for the community and a warning to those using these models today.
Cynthia Rudin - Simpler Machine Learning Models for a Complicated World
Rudin explains why simpler machine learning models often perform better than black box ones (and why we shouldn't necessarily expect the reverse). There's also a mini-concert at the end with songs about machine learning. 'nuff said.
Lee Kezar - Modeling ASL via Linguistic Knowledge Infusion
Kezar shows how different modeling sign language is compared to written languages, the paucity of work in the NLP space here, and introduces some novel, promising methods to move from the standard gesture modeling approach to one that includes linguistic knowledge.
Boris Hanin - Theory of Machine Learning
This is the first of two talks on the list from Hanin, who is quickly becoming one of my favorite theoretical computer scientists. Here Hanin builds up an intuition for the relationship between data, network depth, and network width to understand the limits of very large models, an essential approach for designing these models and evaluating claims about them.
Boris Hanin - Scaling Limits of Neural Networks
This talk gets at fundamental questions in machine learning and provides essential intuition on why/when we can expect models to behave in certain ways - high dimensional optimization efficacy, generalization, feature learning, and appropriately matching data, models, and learning algorithms.
Phillip Isola - The Platonic Representation Hypothesis
Isola explains why large models of various types are all converging to similar representations. I absolutely love this work, and while I think the convergence here is on the world as represented by the Internet rather than the given hypothesis, this approach should fundamentally reshape how companies and researchers value training more models and why phenomena transfer across models.
Michael Saxon - Rigorous measurement for multimodal, multilingual, and multicultural systems
Saxon has done incredible work in this area, and this talk builds on that to demonstrate how to systematically quantify gaps and improvements in these models in important areas.
Dylan Hadfield-Menell - You Can't Have AI Safety Without Inclusion
This talk examines the challenges of specifying goals for algorithms. Hadfield-Menell expertly illustrates why reward systems tend to be brittle due to difficulties in measuring/specifying everything using practical, theoretical, and empirical approaches. He concludes by describing the issue with aligning goals with those of a single individual, advocating strongly for inclusion.
Sanmi Koyejo - The Measure and Mismeasure of AI
Koyejo investigates predictability and surprise in language model benchmarks here, and brings the 🔥, starting with one of the lines of the year: "The new NLP is language models trained by mad libs" 😂, then thoroughly dismantles the notion of "emergence" in these models through rigorous empirical and theoretical work.
Kyle Mahowald - Using Language Models for Linguistics
LLMs are considered to be very good at syntax and grammar, but is this due to memorization, heuristics, or building up a decent model of language? Through a systematic review controlling training data on smaller models, Mahowald convincingly demonstrates that it's the latter, exploring implications for linguistics.
Tina Eliassi-Rad - Just Machine Learning
Eliassi-Rad takes us on a tour of the long history of biased algorithms, ending up at the current day with a refreshingly critical perspective of "red teaming," among other issues.
Sasha Luccioni - Energy Star Ratings for AI Models
Luccioni hits a ton of different topics in this discussion on the climate impacts of large models, including energy consumption, energy star ratings for models, and issues of market concentration, among other topics. My top quote: "Evaluation in AI has become so broken."
Dipto Das - The Colonial Impulse of NLP
Colonial values and biases continue to be perpetuated by sociotechnical systems. Here Das explores these issues by investigating potential bias in sentiment analysis tools in the context of Bengali communities.
Himani Deshpande - Reconfigurable Interfaces by Shape Change and Embedded Magnets
Deshpande demonstrates a reversible shape-changing mechanism that enables reconfigurable 3D printed structures via translations and rotations of parts. They also explore fabrication techniques that enable reconfiguration using magnets and the thermoplasticity of heated polymer.
Céline Coutrix - Impact of Fingernails Length on Mobile Tactile Interaction
Mobile users have fingernails of different lengths, but the impact of different lengths isn’t well understood. Coutrix rectifies that here using interviews and experiments, showing qualitative and quantitative changes in experiences that are important for inclusive device/interaction design.
Lane Lawley - Interactive Task Learning with GPT Dialog Parsing
I HATE when people unthinkingly feed text into LLMs and just assume that whatever comes out is factual/correct. Lawley shows one of the few right ways to use LLMs in areas where the truth matters, developing a carefully scoped system that includes LLMs in proscribed roles, with impressive results.
Elizabeth Stokoe - How "conversational" are conversational technologies?
Given the parrots on the opening slide you know where Stokoe is going, and she brings some absolutely devastating examples on the limitations of these technologies.
Hamsa Bastani - Rethinking Fairness for Human-AI Collaboration
This is some of the best work I've seen in AI in awhile - Bastani emphasizes the importance of optimizing not for the accuracy of an algorithm's prediction in a vacuum, but rather optimizing for the overall system (including the human decision maker). She then demonstrates a general method that does exactly this.
Jonas Geiping - Coercing LLMs to Do and Reveal (Almost) Anything
Geiping goes into detail on important work on developing security exploits for LLMs, the challenges in achieving robustness in large models, and more.
Anthony Chen and Shayne Longpre - The Data Provenance Initiative
This is extremely important work, documenting dataset licensing and attribution in AI development, which is really where the action is in the space.
Simon Lucey - Learning with Less
The focus of this talk is on developing priors for neural networks, and Lucey provides great perspective on recent developments in AI and where these developments have been lacking, presenting compelling evidence that a neural prior approach can help further advance the field.
Sangbae Kim - Embodied Intelligence
Kim systematically goes through how the most successful robotic systems are built today and what they need to be successful, and why the approaches taken to developing generative AI tools are highly likely to not work for building embodied intelligence.
Kim Gallon, Suresh Venkatasubramanian, and Malik Boykin - AI and Power
There's wonderfully no discussion of AI-hype topics here, instead the discussion focuses on the current harms and opportunities created by algorithmic and data intensive technologies. I also loved the criticism of approaches (like synthetic data) that take a techno-solutionist approach to reducing AI harms.
Pedro Lopes - Integrating interactive devices with the user's body
Why build bulky external devices to simulate resistance when you can just as easily use electrical stimulation? Why rig a huge system to blow cold air at a user when you can use computationally controlled emissions of eucalyptol directly to the nasal cavity? You'll learn about all this and more in this fascinating talk.
Daniel Beaglehole - Mechanism of feature learning in neural networks
Beaglehole convincingly demonstrates that the features learned by general neural network architectures are captured by a well understood mathematical operator, significantly increasing our understanding of the uses and limits of these models.
Chinny Sharma - AI's Hippocratic Oath
In this essential talk/discussion, Chinny convincingly lays out the case for professional certification and standards for individual commercial AI developers and draws parallels to certifications in fields like medicine and accounting.
Kyunghyun Cho - Beyond Test Accuracies for Studying Deep Neural Networks
Cho introduces some brilliant experiments that test LLMs for consistency using multiple choice and quotations with devastating results, and in the second half of the talk examines the issues of the same models giving very different solutions depending on a variety of training and tuning factors.
Benjamin Spector - Notes on AI Hardware
This talk gets at the current state of AI hardware, key hardware factors driving the massive performance increases we're seeing in AI, and why it's important for researchers and developers to deeply understand advances and architectures in this space. Spector delivers one of the deepest, clearest dives into some of the leading GPU chips, also identifying areas for possible improvement moving forward.
Lisa Feldman Barrett - Emotion AI: Separating Facts from Fiction
The work presented here systematically deconstructs how computer science and tech more broadly thinks about emotion and the application of AI and other techniques to classifying these fundamentally contextual and experiential concepts.
Ben Zhao - Nightshade: Data Poisoning to Fight Generative AI
Zhao goes over his brilliant work on methods to poison facial recognition and generative image models, which are already out in the wild.
Jay Cunningham - Critical Technology Practices and Social Impact
Cunningham masterfully lays out the importance of viewing and designing technology through a social and ethical lens, complete with devastating examples of the non-neutrality of technology.
Shoko Suzuki - The evolving relationship between humans and technology
While this talk is in Japanese, hopefully auto-captions work because Suzuki’s amazing articulation of critical issues around engaging and shaping AI development literally had me clapping at the end.
Social and Natural Sciences
Sam Zhang - What Science Isn't Being Done?
This talk examines biases and luck in science. Zhang covers so much ground here - scientist career stages, labor inequalities in universities, elite journal biases, and more.
Joe Magee - Are Many Sex/Gender Differences Really Power Differences?
The answer is essentially yes, and there's a hidden meta analysis on the power literature here, in addition to the powerful main effect that there's a 9:1 ratio (!) of consistent to opposite effects.
Michael Tomasello - The Evolution of Agency
Tomasello convincingly demonstrates the different types of agency likely exhibited by our ancestors, how people differ, and the implications for human development.
Tobias Salz - Sources of Market Power in Web Search
Salz presents a rigorous field experiment that changes and measures search provider usage and behavior to tease apart the effects of various features, and even includes surveys!
Raj Chetty - How Children Succeed
Chetty gets into recent work around the surprisingly fast reduction in some inequality metrics recently in the US and implications for policy.
Jean-Baptiste Poline - Statistical and sociological components of reproducibility
Poline explores current problems with replication and reproducibility in the medical sciences, starting at misuse and misunderstanding of statistical methods and touching on poor study planning, hypothesis specification, and incentives.
Russ Poldrack - Reproducibility in fMRI
Poldrack delivers a devastating takedown of the state of neuroimaging research (and indeed all science) today, then examines potential fixes.
Christian Peukert - Strategic Behavior and AI Training Data
Through a rigorous empirical analysis Peukert shows how the release of an online platform's user generated dataset for AI training significantly reduced the quantity, variety, and breadth of future contributions.
Michael Ralph - Insurance, Slavery, and Expertise
Ralph dives into the financial instruments that developed alongside and supported/utilized slavery, with particular focus on how the expertise of enslaved people was valued in insurance contracts and the many implications of these practices for our understanding of US history and current social/financial systems.
Dan Spielman - Algorithmic Discrepancy Theory and Randomized Controlled Trials
Spielman provides deep insight into the challenges in dividing people into groups with nice properties, the surprising effectiveness of simple methods, and his greatest fear - AI proposing a treatment and AI algorithms deciding if it works.
Michael Clemens, Jonathan Portes, and Warwick J. McKibbin - Migration Restrictions and Damage to the US Economy
There are three talks here, and some lively discussion on the topic as a whole:
Clemens - eviscerating unscientific claims about immigration
McKibbin - forecasting Trump’s immigration policies’ negative effects
Portes - immigration, Brexit, and the UK labor market
I learned a ton from Portes' talk about the counterintuitive rapid rise of immigration post-Brexit.
Paul Collier - Left Behind: A New Economics for Neglected Places
Collier clearly identifies how the neoliberal order has accelerated the decline of some geographies and potential fixes. As always, Collier has a banger quote: "Milton Friedman, bless him, it's amazing [that] this guy got a Nobel Prize in economics."
Zeke Hernandez - The Truth About Immigration
Zeke is one of the foremost researchers in this area, and methodically lays out the evidence for the overwhelmingly positive economic and social effects of immigration on the host nation.
Philip Resnik - How Can LLMs Help Us Identify and Use Constructs that We Can Trust?
This is some of the best work I've seen on using LLMs appropriately - essentially as a tool to label - but that usage can be quite powerful.
Daniel Susskind - Growth: a Reckoning
Whenever someone brings up the long-discredited, extremely harmful concept of “degrowth,” I send them this talk. Susskind takes a hammer to this incredibly misguided movement, moving on to the more interesting question of how to rethink growth moving forward.
The Stigler Center’s 2024 Antitrust and Competition Conference, Day 1
The entire event is amazing. The panel on the antitrust history of AT&T and IBM with Robert Crandall, Richard John, Giovanna Massarotto, and Tim Wu was revelatory, as was the following discussion with Randal Picker on the history of antitrust around computers and chips. The panel on litigation against Microsoft with Gary Reback, Ron Schnell, Robert Topel also deserves to be highlighted for excellent perspective.
Matt Beane - Engineering Skill
Matt does a deep dive into how the co-location of low paid workers who helped train robotic systems led to these workers developing other skills and entering more highly compensated career tracks.
Tara Watson - An Economist’s Guide to Immigration Reform
Watson gives a sweeping summary of the overwhelming evidence that immigrants are a huge positive to the economy, even contributing more to government coffers than American-born people. She ends by laying out the case for significantly ramping up legal immigration avenues in the US.
Luigi Zingales and Anat Admati - Straight Talk
Two of the foremost thinkers in economics and business discuss the myth of objectivity in economics, the importance of considering power in research, and so much more.
Övül Sezer - Impression (Mis) Management
Sezer opens with an overview of impression management theory, then delves into fascinating studies about how and why hiding success can backfire.
Mako Hill - Lifecycles of Peer-Produced Knowledge Commons
Mako examines the evolution of different Wikipedia communities, revealing fascinating growth and decline dynamics and relating them to the openness of the platform. He then goes into why absolute openness often becomes undesirable, but the partial closure has some negative effects.
Brian Hare - Survival of the Friendliest
Come for the unique and convincing take on the drivers of human evolution, stay for the bonobo stories.
Pinar Yildirim - Automation, career values, and political preferences
This is an exhaustive look at manufacturing robot introduction in the US from 2000-2016, with Yildrim showing that automation appears to dramatically reduce upward career mobility and expected lifetime earnings. The Q&A session is also not to be missed.
Ethics
Richard Moorhead - Frail Professionalism
Richard expounds on bringing professionalism and ethics back to the practice of laws, reviewing the issues with the current state of legal ethics, particularly in the corporate world, the challenges with improving the status quo, and various paths forward. IMO this kind of reflection is gravely needed in tech and management more broadly.
Josh Cohen - The Reach of Ethics
Cohen explores the types and limits of fairness in AI in this incredible talk. This is one of the best talks I've ever listened to (which longtime watchers know is well over 4K talks at this point), and Cohen articulates points that I've intuitively felt for a while far more clearly than I could hope. This is essential listening.
Sabine Hildebrandt and Howard Israel - The History and Legacy of the Pernkopf Atlas
The Pernkopf Atlas is an anatomical atlas developed by Nazi surgeons, using the bodies of political dissidents, but published in 1960 with its origins meticulously obscured. This tough discussion digs into that history and the ethics of using the Atlas moving forward.
Barbara Prainsack - What is data solidarity and why do we need it?
Prainsack gives a sweeping, compelling account of the need for ethical governance of AI, why individual level control of data isn't effective, and how to build effective regulatory regimes that are informed by a deep understanding of the people's interests and stakes.
Cass Sunstein - The Barbie Problem
Sunstein describes Barbies as "goods that people buy but wish did not exist," and examines social media and other phenomena through this lens. The Q&A also gets more into "reverse Barbies" like paternity leave and the implications for policy.
Eric Mathison - What Should We Do with Coerced People?
What should happen if a patient with full decision-making capacity is coerced? Eric methodically works through this thorny issue, which is especially pressing in the medical context but IMO this approach can greatly inform coerced "consent" in other areas as well (tech, finance, etc.).
Law and Policy
Dina Srinivasan - The End of Google As We Know It?
Srinivasan clearly explains the online ad market that Google is accused of unlawfully dominating, the anticompetitive history behind the current state of the online ad market, and shoots down many common criticisms of the case.
Lauren Pringle - The new DGCL §122 (18) (Stockholder Agreements)
This is a discussion on the recent adoption of DGCL section 122 and the implications for Delaware companies and corporate law more broadly. Pringle clearly lays out the concerns of how this change, which essentially blesses stockholder agreements to supersede board authority, will alter corporate governance moving forward.
Carsten Gerner Beuerle - How to quantify legal rules
This talk explores how to quantify legal rules and the issues with simplified indices in corporate law. Gerner-Beuerle focuses on the problem of interdependence between laws in addition to essentially random heuristics around scoring/measurement, with some suggestions for improving these methodologies.
Diamond Ashiagbor - Race, legal form, and the labor contract
Ashiagbor shows how the law was changed to allow for slavery across the British colonies, and how even abolition didn't remove some of the legal changes that supported the institution of slavery and its descendants.
James Grimmelman - Does Generative AI Infringe Copyright?
Grimmelmann provides a good overview of the issues at play here, as well as a savage dig at Gerald Ford😲 Favorite quote: "It's not at all obvious that the incentives to create of the sort that copyright offers are the appropriate system of law to govern this new [technology]. It may be that what replaces copyright due to generative AI is as different from copyright as copyright was from the patronage system that came before it."
Ann Lipton - Value vs. Values: ESG Investing and the Social Responsibility of Business
Lipton is one of the top thinkers in this area and it's not hard to see why - starting with a fantastic look at the US history of the corporate form, she then clearly summarizes the shareholder primacy vs. stakeholder theory debate, and finally reviews ESG investing and how current political headwinds make confronting the need to reform corporate law and policy inevitable. I have some small quibbles around the edges here, but this is a can't miss talk.
Ryan Calo - Law and Technology in Theory and Practice
Calo details the various legal hurdles to the successful legal analysis and governance of emerging technology, and also cites the relevant work of my late cousin Joel Reidenberg!
Oren Bracha - Physicalism in Intellectual Property
Bracha delivers deep insight about how physicalism seeps into many aspects of current IP law and jurisprudence, ending with an examination of issues around generative AI in this space.
Daniel Solove - Murky Consent
Solove rightly points out the fiction of consent, and why embracing that fiction can lead to a more effective privacy law.
Martin Sybblis - Decolonization and Corporate Law
This is a fascinating examination of the differential economic and social outcomes of different Caribbean countries and the role that corporate law (particularly tax law) played and continues to play in the decolonization process.
Isabella Weber, Jan Eeckout, Gabriel Zucman, Doha Mekki, and Florian Ederer - Market Power Dysfunctions: Seller Inflation, Inequality and AI Acceleration
This is an amazing panel on market power dysfunctions and the issues with traditional economic approaches. Florian has the line of the conference that this panel is from: "The gospel of Berry-Levinsohn-Pakes says we must use the holy ritual of demand estimation and use the holy purification ritual of Hausman instruments."
Andreas Mundt, Luigi Zingales, Tommaso Valletti, Becca Kelly Slaughter, and Gina Cass-Gottlieb - “From Price to Power”? Reorienting Antitrust for the New Political Economy
This is an all-star panel on reorienting antitrust around power analyses. There's so much packed into this conversation around the importance of antitrust, the evolving nature of corporate power and antitrust policies, and more.
Eugene Volokh - Large Libel Models? Liability for AI Output
Volokh systematically lays out the convincing case for LLM vendor liability for model outputs, also explaining why open source models likely fall under a different regime.
Shareholder Primacy
Ann Lipton and Michael Levin have put together one of the best podcasts out there, examining a variety of late-breaking corporate governance issues with engaging wit and insightful analysis. It’s too hard for me to choose a few of their shows from this year for the list, so just listen to them all.
Books (in no particular order)
Virginia Eubanks - Automating Inequality
Virginia Eubanks lays out a powerful, deeply researched case for reconsidering the role of technology and algorithms in vital human services. By considering in depth a number of specific cases of how different technologies have created Kafka-esque systems that ossify and amplify existing societal biases, Eubanks challenges technologists, researchers, and governments to rethink how and what they build. This book also contains devastating historical analyses of decision making relating to services for the poor, showing that today's efforts are not an innovation or an aberration.
Sven Beckert - Empire of Cotton
Beckert uses the developments of the cotton industry - from small scale production based in the Ottoman Empire and India disseminated through trade networks to slavery and colonialism-fueled mechanized production to the modern industry - to examine the evolution of state power and capital-driven private industry through the centuries. The bulk of this book focuses on the period immediately prior to the industrial revolution until the end of 1800s, and for good reason - this was arguably one of the most important eras in history through the lens of industrial and political realignment. Beckert illustrates how modern supply chains, futures markets, and waged labor all emerged from this turbulent period, as well as how many of the inequities that were imposed on different groups and regions are a direct result of the relentless push of states and industrialists to extract ever more value out of this fantastically profitable enterprise.
I particularly liked how this book combines history with a statistical accounting of changes over time, providing a much more concrete sense of the scale of economic and political changes the world over. This book could certainly be longer - the modern clothing and cotton industry is scarcely examined - but given that this work is concerned with many more unique innovations of business and politics (and the modern clothing industry is probably more similar today to a number of other large scale manufacturing industries than it was in the past) that's a reasonable tradeoff.
Overall, this book is a must read if you're at all interested in global economic history.
Claudio Saunt - Unworthy Republic
Claudio Saunt narrows the focus of Native American dispossession in modern day Georgia and Florida during the Andrew Jackson administration, and that focus provides a harrowing account of a racially-driven dispossession machinery in unprecedented detail. It's hard to read some of these accounts and not see echoes of them in Nazism and other bureaucracy-driven genocides in the following years - the huge administration that was stood up to deport Native Americans, the strict accounting processes and budgets, and the dehumanization of an entire people.
Saunt deftly mixes first hand written accounts, historical records, and statistics to provide a holistic perspective on this tragedy. The juxtaposition of atrocities with current cities and attractions in Georgia and Florida is jarring, with the lack of American acknowledgement made that much more difficult to comprehend. While this is a tough read, it should be required reading for those who want to better understand a formative part of American and world history and gain perspective on current attempts to repeat it.
Ruha Benjamin - Race After Technology
I've heard many talks by Ruha Benjamin and others on some of the critical themes covered in this book: how facially "neutral" designs can ossify and expand inequity, how attempts to reduce bias can in fact increase it, etc. I've even heard many of the examples in this book before: the shockingly racist systems embedded in film development technology, the replication of eugenics throughout the decades, and so on. But this book covers each topic with a rigor and depth that is so illuminating, so insightful, that it demands to be read if you're even peripherally connected or use technology (basically everyone).
Benjamin delivers a clarion call for thinking through the systems in which we embed technology, the need to question what is measured and optimized, and the importance of building or dismantling systems to bring about a more just world.
Caitlin Rosenthal - Accounting for Slavery
Most management textbooks start with a review of "scientific management," but Rosenthal demonstrates why scholars should look back farther to the slave plantations of the 18th and 19th century for the genesis of modern approaches to accounting and management. Using volumes of historical records, this book shows how the slave plantation industry developed sophisticated methods to control and measure every aspect of their plantation, including their slaves. There are direct lines from these practices to the development of org charts and time and motion studies - Henry Gantt of Gantt chart fame, for example, grew up in a family that had grown rich from owning slaves and almost certainly used these same management methods.
All of this cries out for a reckoning with different management practices that have become commonplace - individual work measurement, the myopic focus on easy to measure quantitative metrics, etc. If you're in management or people analytics even peripherally, this is a must read.
Katharina Pistor - Code of Capital
Most people, including some of the world's most prominent economists, believe that capital is some natural phenomena. That land ownership is simply a given, that inventions are your property, and that company shareholders have no liability for the company's failings or misdeeds. Katharina Pistor takes those assumptions to the woodshed in this tour de force, winding through the historical choices of rulers, lawyers, and merchants to encode assets as legal entities. These forces, far from being natural, are shaped by societal choices, and can be unmade or modified in similar ways.
This book vividly illustrates how those with power have used old legal instruments and new legal innovations to protect their wealth and the implications of these protections on perpetuating and accelerating inequality. The historical developments of the enclosure of the commons are particularly vivid, as is the opening example of the Belize government's appropriation of Mayan land. The centrality of private attorneys in driving asset encoding is also illuminating, particularly in light of many recent legal changes in the US to privilege that kind of law creation. While there is a slight detour into blockchain that is best sped through (the book was published in 2020, after all), overall this book gives a fantastic overview of legal mechanisms of capital creation and protection and provides a new lens through which to view the modern economy.
Michael Tomasello - Becoming Human
As humans, we like to think that we're radically different from other animals. When it comes to our closest living relatives, however, it is often quite challenging to articulate our precise differences. Tomasello dives into the vast pool of fascinating experimental academic research in this area, taking us through stages of human and ape cognitive, social, and cultural development through childhood to illuminate uniquely human traits and their importance is our rise to global dominance.
Tomasello deftly marries readability with academic rigor, which I've found to be extremely rare. It doesn't hurt that the experiments themselves are ingenious and often accompanied by hilarious quotes from small children. If you're at all interested in child development, evolution, or cognition, this is a must read.
Sara Hendren - What Can a Body Do?
Humans are inextricably connected with our tools - from the explicit (can openers) to the unnoticed (t-shirts) to the systemic (building architecture and urban design). This book investigates who these tools are designed for, who designs them, how they're designed and evaluated, and what happens when we expand the aperture of those current answers. Hendren viscerally demonstrates the power and importance of focusing on the mismatches that particular individuals face and meeting them with design. Rather than aiming for products that scale uniformly across large masses of people, this approach advocates for developing and scaling design processes.
This book is an engaging mix of personal anecdotes and rigorous academic research and theory. The historical background on some of the issues examined in this book, while necessarily brief, still provides tantalizing perspectives that interested readers can follow up on. This is a vital, one of a kind book.
Kim TallBear - Native American DNA
Kim TallBear delivers a tour de force, eviscerating the "objectivity" of genetic testing (and science more broadly) and situating that mindset as it approaches the application of DNA testing to Native American populations. The careful consideration of what being a tribal member means, how it's fundamentally unrelated to genetics except in extremely limited situations, and how researchers have systematically recreated eugenic, racist categorizations (often unintentionally) is impressive and devastating. Importantly, this examination applies to DNA testing approaches for non-Native Americans as well - it just is fundamentally unequipped to answer questions of individual ancestry.
For anyone interested in the ethics and science of genetic testing this book is absolutely essential, but I would argue that anyone dealing with data analysis more broadly would benefit immensely from this book.
Dorothy Roberts - Fatal Invention
Attempts to identify biological roots of race, an inherently socially defined phenomena, goes back centuries. Roberts traces those roots here, leading to modern genetic science and fundamentally flawed attempts to use quantitative methods to claim objectivity. This book systematically deconstructs those attempts, identifying the real harms that they cause and charts paths forward for genetic science that leaves the concept of race to the social sciences.
Unfortunately this book still resonates deeply, despite the fact that it was written in the middle of the Obama administration. Roberts is incredibly prescient here, hypothesizing that the current framework of race-driven genetics will further deepen existing systematically racist systems. This work is important not just for those in the biological sciences, but folks in management and technology that use analogous methods.
David Treuer - The Heartbeat of Wounded Knee
David Treuer presents, in this moving and insightful and deeply researched book, a holistic perspective on the reality of modern Native Americans. Combining historical research, qualitative interviews, and some quantitative methods in the later sections, one gets a sense of the continuing evolution of this essential part of American society.
There is still decent space given to level set on the past, covering Native American history prior to colonization and prior to 1890. After this section Treuer goes through detailed accounts of the shameful and fraught interactions between tribes and the US government, highlighting some successful negotiations and others that ended with further appropriation of native lands. There are brief glimpses of more enlightened policy makers, with some laws and organizational decisions that continue to reverberate today.
The last few parts of the book are nothing short of stunning. The heartfelt study of the activist period of the 60s and 70s was profound, and the legal challenges that led to some tribes embracing gaming and further sovereign rights was illuminating. The sober analysis of the economic challenges many Native Americans face is also contrasted with some incredible success stories. Those success stories leave one wanting a full economic analysis of their underlying causes, but given Treuer's academic background it's probably best that he just set the table here for other scholars.
Especially if you're an American like myself, you'd be hard pressed not to find yourself moved and motivated by this book.
Joshua Specht - Red Meat Republic
This book presents a powerful case for the US beef industry as a microcosm of changes wrought by the Industrial Revolution in the US. Starting with the dispossession of Native Americans and their role in the bison industry, Specht charts how the spread of the railroads allowed for radical shifts in beef consumption. Paired with refrigeration technology and global capital flows, he shows how this accelerated the delivery of beef to ever wider numbers of people and how it facilitated a concentration of wealth and power in chokepoints of this industry - the nascent commodity traders and meat packers.
This book also covers how the period of the "cowboy" and the myth of widespread beef eating began, with essentially no basis in reality given that large scale cattle farming didn't exist before the railroads. The familiar beat of US consumers protesting high beef prices but not working conditions in the industry, the desire to return to a "natural" diet, and the increasingly gendered notions of meat consumption demonstrate how timeless these trends are and how little progress has been made in the 100+ years since the manifestation of the large-scale been industry.
Jonathan Haskel and Stian Westlake - Capitalism Without Capital
Most of the time we don't think about how companies and governments account for investments, assets, and costs, but it is essential for our understanding of the state of these entities and how they're changing over time. Haskel and Westlake detail the gradual shift from easily accountable, tangible production making up the bulk of economic growth to today's world where investments in processes, knowledge, and intellectual property make up the majority. In doing so they illustrate why this difference is critical for everything from investment in and prioritization of government policies to investment strategies and vehicles to corporate strategies.
There are a few hiccups in this journey, particularly in the cursory and somewhat off treatment of topics better categorized as organizational behavior and urban planning (and the anecdotal takes of some demonstrably misguided and failing entrepreneurs should be relegated to the dustbin). Still, those sections don't take the shine off a phenomenal book on a critical topic.
Anthony Abraham Jack - The Privileged Poor
Elite college access can certainly be extremely beneficial for students from poor backgrounds, but barriers to the full benefits of college once students are admitted is criminally understudied. This book rectifies that gaping hole in the literature through cutting interviews and supporting research, detailing the significant differences between the "privileged poor" students and the "doubly disadvantaged" students, and points to a number of areas where colleges can relatively easily and cheaply significantly improve to create a more just and equitable environment.
The appendix is also excellent, detailing the research methodology used for the book and best practices for this kind of work.
David Graeber and David Wengrow - The Dawn of Everything
Graeber and Wengrow have seemingly done the impossible: upend centuries of accepted Western thought about the development of human societies and the genesis of the last three centuries of political thought. Using exceptionally thorough research, this book demonstrates how it is exceedingly likely that human societies have continually cycled through egalitarian and hierarchical regimes, sometimes even in the same year, and how the exact configuration a society found itself in is by no means a deterministic function of societal scale or complexity. Beyond that, they convincingly shred earlier dogma around the sudden appearance of agriculture, showing how small scale seasonal gardens (like we have today) existed and even outcompeted larger scale agricultural practices. Finally, through rigorous archival work they make a strong case for enlightenment thought originating in the indigenous people of North America rather than Europe.
All of this is accomplished with an accessible and engaging style that deftly disarms criticisms by turning them against themselves with straightforward logic. The result is nothing short of a revolutionary treatise on human society and what we are capable of when political possibilities are more open.
Tom Standage - The Victorian Internet
Technology interacts with society in unpredictable ways, and arguably one of the first instantiations of a technological revolution that upended norms, business models, and the law was the telegraph. As Standage shows in this masterful book, the advent of the telegraph spawned changes that are shockingly similar to those we've observed with the internet and other modern technologies - utopian prognostications about its effects, new forms of military action, platform business models, network effects, disinformation, and more.
Our distance from the time of the telegraph makes it particularly instructive, and one that economists and technologists alike would be wise to reflect on when pontificating on the unique nature or general applications of a particular new innovation.
Augustine Sedgewick - Coffeeland
Using the development of El Salvador's coffee industry as a jumping off point, Sedgewick convincingly demonstrates how coffee production is inextricably linked to colonialism, technological development, and accidents of history. Sedgewick brings deep research and a holistic approach to this instructive topic, and his extremely engaging writing style makes the entire book that much more enjoyable.
The importance of scientific advances in the mid-1800s was particularly interesting to me, with the discovery of the law of thermodynamics and the fascination with measurement spilling over to the workplace and society more broadly. The examination of its role in facilitating the expansion of scientific management and caloric expenditure measurement, first in the lab and then put in practice on plantations through an elaborate cafeteria and food production system, echoes even today in free corporate lunches. The importance of the human choices made during this period that dramatically shaped working conditions for plantation workers and the growth of the coffee industry was shocking and incredible.
Overall, this book not only highlights the genesis of the industry that produces a staple of the global diet. It also underlines how economic demand and concern for profit alone can create horrible societal outcomes that are hidden from consumer markets.
Lars Chittka - The Mind of a Bee
Lars Chittka has penned a masterpiece, using witty writing with fascinating narratives around the genesis of different experiments to lay out a solid case for bees as capable of far more cognition than most people think. Starting from an exploration of bee sensory mechanisms (the sections on navigation and light polarization were amazing), then moving to a wide variety of experiments to detail just how much bees can plan, remember, and reason. Some of the sections on group behavior were less compelling, since while Chittka briefly explores complexity theory it wasn't touched on in the section on hive construction.
Overall this is one of the best books on cognition that I've read, and with all the hype about "general intelligence" folks would be well advised to pick up a copy of this book before opining on the topic.
Eduardo Bonilla-Silva - Racism Without Racists
This book is rigorous, honest, devastating, vital, and (unfortunately) still timely. Combining macro sociological analysis, in-depth interviews, and a biting wit, Bonilla-Silva has produced a masterpiece that lays bare the racism that still lies at the heart of American society. The updates of this book after its initial publication are even more illuminating, continually reinforcing the points made in earlier analyses. I started listening to the audiobook version of this with my kids, and it's one of those rare books that has to be absorbed by every generation, and I would argue everyone in society.
Gina Rippon - The Gendered Brain
Combining neuroscience and psychology, Gina Rippon delivers a systematic takedown of the "female brain" myth as well as intrinsic sex differences in general. It might be more accurate to call this book "the gendered individual" given how much psychology is brought in here. This is necessary, however, as Rippon demonstrates how quickly societal and caregiver expectations influence behavior and activities.
One of the challenges with examining brains for differences between groups is that brains aren't static. As one grows and experiences and interacts with the world, they are influenced and changed. As such, the broad approach Rippon takes here is necessary. Even if some of the psychology studies she cites haven't replicated well, she takes pains to note study limitations, and the statistical rigor and deconstruction of poor studies is refreshing.
Katharina Pistor and Curtis Milhaupt - Law & Capitalism
Pistor and Milhaupt have put together a masterful, carefully researched book that thoroughly refutes the notion of the law as an exogenous, neutral factor in economic models, definitively showing that there is a rolling relation between law and markets that involves local constituencies (management, investors, the public, employees, politicians, etc.) mediating their relationship. Mercifully they go well beyond the standard US and Western Europe lens, bringing in Japan, Korea, China, and Singapore (among others) in fantastic detail.
The case study method is employed throughout this book, which in addition to providing insight into recent corporate scandals vividly illustrates the limits of some of the existing structures that underlie corporate governance in different jurisdictions. They also show that the common vs. civil law dichotomy, as well as the various benchmarks that are based on particular items of law, offers very little in terms of understanding why certain economies are successful. Instead they argue for viewing legal systems along the centralized-decentralized and coordinative-protective axes, which dovetails nicely with their case analyses. Overall, they show how varied the legal and regulatory regimes are around the world that have all led to economic success, expanding the lens of what good corporate governance legal regimes can look like.
Connecting people, problems and institutions. Working on improving data quality.
2wSo much fun. Thx
Thanks to all the book authors: Virginia Eubanks, Sven Beckert, Claudio Saunt, Ruha Benjamin, Caitlin Rosenthal, Katharina Pistor, Sara Hendren, Kim TallBear, Dorothy Roberts, David Treuer, Joshua Specht, Jonathan Haskel, Stian Westlake, Anthony Jack, David Graeber, David Wengrow, Tom Standage, Augustine Sedgewick, Lars Chittka, Eduardo Bonilla-Silva, Gina Rippon, and Curtis Milhaupt
Thanks to all the speakers in the corporate law and policy area: Dina H. Srinivasan, Lauren Pringle, Carsten Gerner-Beuerle, Diamond Ashiagbor, James Grimmelmann, Ryan Calo, Oren Bracha, Daniel Solove, Martin Sybblis, Isabella Weber, Jan Eeckout, Gabriel ZUCMAN, Doha Mekki, Florian Ederer, Andreas Mundt, Tommaso Valletti, Becca Kelly Slaughter, Gina Cass-Gottlieb, Eugene Volokh, Ann Lipton, and Michael Levin
Thanks to all the speakers in the ethics area: Richard Moorhead, Josh Cohen, Sabine Hildebrandt, Howard Israel, Barbara Prainsack, Cass Sunstein, and Eric Mathison, PhD, HEC-C