What do you think?
Rate this book
Fatal Abstraction: Why the Managerial Class Loses Control of Software
A tech insider explains how capitalism and software development make for such a dangerous mix.
Software was supposed to radically improve society. Outdated mechanical systems would be easily replaced; programs like PowerPoint would make information flow more freely; social media platforms like Facebook would bring people together; and generative AI would solve the world’s greatest ills. Yet in practice, few of the systems we looked to with such high hopes have lived up to their fundamental mandate. In fact, in too many cases they’ve made things worse, exposing us to immense risk at the societal and the individual levels. How did we get to this point?
In Fatal Abstraction, Darryl Campbell shows that the problem is “managerial software”: programs created and overseen not by engineers but by professional managers with only the most superficial knowledge of technology itself. The managerial ethos dominates the modern tech industry, from its globe-spanning giants all the way down to its trendy startups. It demands that corporate leaders should be specialists in business rather than experts in their company’s field; that they manage their companies exclusively through the abstractions of finance; and that profit margins must take priority over developing a quality product that is safe for the consumer and beneficial for society. These corporations rush the development process and package cheap, unproven, potentially dangerous software inside sleek and shiny new devices. As Campbell demonstrates, the problem with software is distinct from that of other consumer products, because of how quickly it can scale to the dimensions of the world itself, and because its inner workings resist the efforts of many professional managers to understand it with their limited technical background.
A former tech worker himself, Campbell shows how managerial software fails, and when it does what sorts of disastrous consequences ensue, from the Boeing 737 MAX crashes to a deadly self-driving car to PowerPoint propaganda, and beyond. Yet just because the tech industry is currently breaking its core promise does not mean the industry cannot change, or that the risks posed by managerial software should necessarily persist into the future. Campbell argues that the solution is tech workers with actual expertise establishing industry-wide principles of ethics and safety that corporations would be forced to follow. Fatal Abstraction is a stirring rebuke of the tech industry’s current managerial excesses, and also a hopeful glimpse of what a world shaped by good software can offer.
Software was supposed to radically improve society. Outdated mechanical systems would be easily replaced; programs like PowerPoint would make information flow more freely; social media platforms like Facebook would bring people together; and generative AI would solve the world’s greatest ills. Yet in practice, few of the systems we looked to with such high hopes have lived up to their fundamental mandate. In fact, in too many cases they’ve made things worse, exposing us to immense risk at the societal and the individual levels. How did we get to this point?
In Fatal Abstraction, Darryl Campbell shows that the problem is “managerial software”: programs created and overseen not by engineers but by professional managers with only the most superficial knowledge of technology itself. The managerial ethos dominates the modern tech industry, from its globe-spanning giants all the way down to its trendy startups. It demands that corporate leaders should be specialists in business rather than experts in their company’s field; that they manage their companies exclusively through the abstractions of finance; and that profit margins must take priority over developing a quality product that is safe for the consumer and beneficial for society. These corporations rush the development process and package cheap, unproven, potentially dangerous software inside sleek and shiny new devices. As Campbell demonstrates, the problem with software is distinct from that of other consumer products, because of how quickly it can scale to the dimensions of the world itself, and because its inner workings resist the efforts of many professional managers to understand it with their limited technical background.
A former tech worker himself, Campbell shows how managerial software fails, and when it does what sorts of disastrous consequences ensue, from the Boeing 737 MAX crashes to a deadly self-driving car to PowerPoint propaganda, and beyond. Yet just because the tech industry is currently breaking its core promise does not mean the industry cannot change, or that the risks posed by managerial software should necessarily persist into the future. Campbell argues that the solution is tech workers with actual expertise establishing industry-wide principles of ethics and safety that corporations would be forced to follow. Fatal Abstraction is a stirring rebuke of the tech industry’s current managerial excesses, and also a hopeful glimpse of what a world shaped by good software can offer.
315 pages, Kindle Edition
Published April 8, 2025
About the author
Darryl Campbell
3 books5 followersDarryl Campbell has worked at Amazon, Uber, Expedia, and Tinder. His writing has appeared in the Verge, Vulture, and GQ. He lives in Dallas, Texas.
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 15 of 15 reviews
October 14, 2025
As somebody who has worked in tech for over 15 years at this point, I've seen many of the same dynamics that Campbell ably puts down in this book. The main thesis is that there is a class of executives who all have been trained at MBA school to look at all business as a pure abstraction of the same metrics or KPIs, no matter what the industry or product. This has been a shift from where engineers and people who actually had knowledge of the product-development process had power to push back on ideas that seemed fine at a high-level, but that can cause extremely harmful outcomes when it actually goes to market.
Sometime this isn't a big deal, but, in the case of Boeing airplanes, the shift to cut costs, increase ROI, sacrifice quality for speed, and to squeeze vendors led to a flawed product design that killed hundreds of passengers. Campbell is at his best in his chapter that focuses on how this design came about, the logic behind it, and why it was fatally flawed. He's able to show other examples from companies like Uber killing a pedestrian with a self-driving car to our destructive new Generative AI collective mania to show how "managerialism" has infected Capitalism. A good, topical book that I'd recommend.
Sometime this isn't a big deal, but, in the case of Boeing airplanes, the shift to cut costs, increase ROI, sacrifice quality for speed, and to squeeze vendors led to a flawed product design that killed hundreds of passengers. Campbell is at his best in his chapter that focuses on how this design came about, the logic behind it, and why it was fatally flawed. He's able to show other examples from companies like Uber killing a pedestrian with a self-driving car to our destructive new Generative AI collective mania to show how "managerialism" has infected Capitalism. A good, topical book that I'd recommend.
July 27, 2025
This book was deeply bleak and depressing, and made me feel vaguely sick as I read it. Although the tour through software failures throughout the years was interesting, the argument made throughout the book was not cogent, specific, or novel. The call to action to tech workers alone, with the author throwing up his hands at regulation or transformative change as hopeless, was laughable, especially as it did not offer any specifics about tech unionization. Reading the acknowledgements uncovers the fact that this was a solid story about the Boeing 737 MAX crashes that was stretched beyond its limits to make a vague argument most of its readers will probably already just as vaguely agree with.
December 7, 2025
In 2025's Fatal Abstraction, Darryl Campbell leverages his years of experience working at prominent tech companies (including Uber, Amazon and Tinder) to criticize the structure of tech or tech-adjacent companies in over-reliance on software instead of frontline workers'/engineers' expertise.
This is an interesting, albeit not particularly novel, viewpoint (see my further reading recommendations below). Campbell's strongest argument is about the Boeing 737 MAX debacle, where software overpowering human pilot expertise led to several fatal crashes (a topic about which he's already published many articles), and he makes a strong case comparing that to the older Airbus A320 where software and human ingenuity in times of crisis are more balanced (i.e., the miracle on the Hudson).
Campbell's other arguments vary from reasonably strong to less relevant. I found his argument about how the Columbia shuttle disaster (where the space shuttle disintegrated upon reentry due to lack of insulation foam that was damaged during takeoff) being partly the fault of Powerpoint to be particularly weak. Campbell insinuates that because the work group who computationally modeled the effects of foam loss were only allotted a single Powerpoint slide to present during a NASA briefing before making the decision to allow reentry to proceed, the constraints of Powerpoint were partially to blame for the shuttle's loss. I presume that might be true if someone's presentation skills are so bad that they're only capable of reading word-for-word off a Powerpoint slide and don't have the foresight to strategically design the slide with only the most salient figures and words, rather than using their brain and their own words to augment the slide's arguments and focus audience attention appropriately. Campbell touches a bit on AI/ML though surprisingly that's not the main focus of this book (though there are plenty of books coming out every week on this very topic, oversaturating the market).
This genre of books almost always ends with the author being asked to propose solutions for the problems they spend the bulk of the book explaining, and this book is no exception. Campbell's proposed solutions are that tech workers should unionize and insert more checks and balances into the system -- i.e., less unbridled power by the MBAs and executive managerial showrunners. This is a nice thought, but difficult in theory. What's ideal (though not always realistic) is to cross-train tech workers in management and financial skills, so that ultimately tech companies are helmed by people with all necessary skillsets. We see that at companies like Jensen Huang's Nvidia (see Tae Kim's The Nvidia Way). I work in healthcare and I see the same argument all the time -- the folks with the expertise and burden of providing the healthcare are beholden to C-suite decision makers with MBAs and finance backgrounds who make decisions chasing a bottom line but totally ignorant of what's actually happening on the hospital wards.
Further reading: existential quandaries of big tech
Race After Technology: Abolitionist Tools for the New Jim Code by Ruha Benjamin
The Big Nine: How the Tech Titans and Their Thinking Machines Could Warp Humanity by Amy Webb
Code Dependent: Living in the Shadow of AI by Madhumita Murgia
My statistics:
Book 362 for 2025
Book 2288 cumulatively
This is an interesting, albeit not particularly novel, viewpoint (see my further reading recommendations below). Campbell's strongest argument is about the Boeing 737 MAX debacle, where software overpowering human pilot expertise led to several fatal crashes (a topic about which he's already published many articles), and he makes a strong case comparing that to the older Airbus A320 where software and human ingenuity in times of crisis are more balanced (i.e., the miracle on the Hudson).
Campbell's other arguments vary from reasonably strong to less relevant. I found his argument about how the Columbia shuttle disaster (where the space shuttle disintegrated upon reentry due to lack of insulation foam that was damaged during takeoff) being partly the fault of Powerpoint to be particularly weak. Campbell insinuates that because the work group who computationally modeled the effects of foam loss were only allotted a single Powerpoint slide to present during a NASA briefing before making the decision to allow reentry to proceed, the constraints of Powerpoint were partially to blame for the shuttle's loss. I presume that might be true if someone's presentation skills are so bad that they're only capable of reading word-for-word off a Powerpoint slide and don't have the foresight to strategically design the slide with only the most salient figures and words, rather than using their brain and their own words to augment the slide's arguments and focus audience attention appropriately. Campbell touches a bit on AI/ML though surprisingly that's not the main focus of this book (though there are plenty of books coming out every week on this very topic, oversaturating the market).
This genre of books almost always ends with the author being asked to propose solutions for the problems they spend the bulk of the book explaining, and this book is no exception. Campbell's proposed solutions are that tech workers should unionize and insert more checks and balances into the system -- i.e., less unbridled power by the MBAs and executive managerial showrunners. This is a nice thought, but difficult in theory. What's ideal (though not always realistic) is to cross-train tech workers in management and financial skills, so that ultimately tech companies are helmed by people with all necessary skillsets. We see that at companies like Jensen Huang's Nvidia (see Tae Kim's The Nvidia Way). I work in healthcare and I see the same argument all the time -- the folks with the expertise and burden of providing the healthcare are beholden to C-suite decision makers with MBAs and finance backgrounds who make decisions chasing a bottom line but totally ignorant of what's actually happening on the hospital wards.
Further reading: existential quandaries of big tech
Race After Technology: Abolitionist Tools for the New Jim Code by Ruha Benjamin
The Big Nine: How the Tech Titans and Their Thinking Machines Could Warp Humanity by Amy Webb
Code Dependent: Living in the Shadow of AI by Madhumita Murgia
My statistics:
Book 362 for 2025
Book 2288 cumulatively
August 1, 2025
I am ten pages in and struggling to choke down the regurgitated ‘us vs. them’ rhetoric of a teenager talking about their chores while plowing through a first person shooter after school.
Early (ten pages in early), the author stakes a claim that ‘managerialism’ has created a class of people who know nothing about what their employees do, and while telling a story about software, cites the catastrophic use of DDT by Montsanto as an example of managerialism in tech.
He closes this narrative by describing manager as [sic] ‘dogs trying to understand physics’. Surely you can’t teach physics to dogs, and that makes managers bad!
I love dogs, I’ve written more code than the author, and yes, I’m manager in tech. You can discount my review, but the author knows nothing about writing good software or managing people. He has written a book wrapped around the bubble gum of a sophomore year in the study or ‘people in charge suck’.
Even then, he makes as much of a case for the work he disdains in his opening paragraphs:
After ww2x ‘corporate profits had increased more than tenfold’, and says that there is a ‘strong positive relationship between the adoption of managerial practices and the success of a business’. But we know he’s angry, and we know he wants us to keep reading because managers are dumb, and physics is hard, and you can’t teach physics to dumb people, especially when the dumb people are running a for profit company.
I hope you save your time.
Early (ten pages in early), the author stakes a claim that ‘managerialism’ has created a class of people who know nothing about what their employees do, and while telling a story about software, cites the catastrophic use of DDT by Montsanto as an example of managerialism in tech.
He closes this narrative by describing manager as [sic] ‘dogs trying to understand physics’. Surely you can’t teach physics to dogs, and that makes managers bad!
I love dogs, I’ve written more code than the author, and yes, I’m manager in tech. You can discount my review, but the author knows nothing about writing good software or managing people. He has written a book wrapped around the bubble gum of a sophomore year in the study or ‘people in charge suck’.
Even then, he makes as much of a case for the work he disdains in his opening paragraphs:
After ww2x ‘corporate profits had increased more than tenfold’, and says that there is a ‘strong positive relationship between the adoption of managerial practices and the success of a business’. But we know he’s angry, and we know he wants us to keep reading because managers are dumb, and physics is hard, and you can’t teach physics to dumb people, especially when the dumb people are running a for profit company.
I hope you save your time.
October 20, 2025
Fatal Abstraction: Why the Managerial Class Loses Control of Software by Darryl Campbell is a riveting, incisive exploration of how corporate ideology and code collided to shape and endanger the modern world.
Campbell, a rare voice who moves fluently between technology, philosophy, and cultural critique, reveals a disturbing truth: the crisis in tech isn’t born from innovation, but from abstraction a mindset that replaces understanding with metrics, people with systems, and responsibility with optimization. Through vivid analysis spanning airline failures, algorithmic bias, and AI hubris, he exposes how the managerial elite’s obsession with control has created the very chaos it sought to prevent.
What emerges is both a diagnosis and a warning that software doesn’t merely automate; it amplifies human arrogance, especially when governed by those who neither build nor fully comprehend it. Campbell’s argument is as philosophical as it is urgent, positioning developers not as passive implementers, but as moral agents capable of reining in the machine of capitalism itself.
With its blend of investigative clarity and intellectual gravitas, Fatal Abstraction stands alongside works like Shoshana Zuboff’s The Age of Surveillance Capitalism and Tristan Harris’s The Social Dilemma a must-read for anyone who believes the next revolution won’t be digital, but ethical.
Campbell, a rare voice who moves fluently between technology, philosophy, and cultural critique, reveals a disturbing truth: the crisis in tech isn’t born from innovation, but from abstraction a mindset that replaces understanding with metrics, people with systems, and responsibility with optimization. Through vivid analysis spanning airline failures, algorithmic bias, and AI hubris, he exposes how the managerial elite’s obsession with control has created the very chaos it sought to prevent.
What emerges is both a diagnosis and a warning that software doesn’t merely automate; it amplifies human arrogance, especially when governed by those who neither build nor fully comprehend it. Campbell’s argument is as philosophical as it is urgent, positioning developers not as passive implementers, but as moral agents capable of reining in the machine of capitalism itself.
With its blend of investigative clarity and intellectual gravitas, Fatal Abstraction stands alongside works like Shoshana Zuboff’s The Age of Surveillance Capitalism and Tristan Harris’s The Social Dilemma a must-read for anyone who believes the next revolution won’t be digital, but ethical.
April 22, 2025
Part of a wave of books that are a backlash against the excesses of the Silicon Valley growth-at-all-costs tech mindset. Highly recommended as a cautionary tale as well as a manifesto for how things can get better (hint: tech workers unionize). Weaves together the connection between airplane crashes, social media, powerpoint, AI-at-all-costs.
June 18, 2025
An extremely insightful perspective of the tech industry spoken from first hand encounters of managerial failures. These experiences speak the true story of managerialism in our modern age and offer a cautionary tale of accepting technology outright.
March 18, 2026
Rife with cliches about software and the industry. Would have been more interesting to go into the economics of software business and why they get overrun by a managerial class that’s obsessed with horizontal, commoditized software that doesn’t actually move the needle
May 17, 2025
Interesting view of a company's view of software. Good read.
July 22, 2025
Excellent work for understanding how software interacts with and shapes modern life and the pitfalls of managerialism when applied to new technologies.
August 5, 2025
[ Review to come: this might require a serious review on my blog. ]
August 31, 2025
His central premise is good: use software to augment humans and don’t focus only on the financial bottom line. But this book is bad and his arguments are stupid.
November 21, 2025
Great! Super scary and invoked sheer terror about the world, but great!
February 16, 2026
4.5 - if you know someone who recently left a high paying tech job and/or the industry entirely and want to understand why, read this.
July 15, 2025
Fascinating but a dense read. Kinda lost me towards the end. Most memorable section to me was regarding Boeing. Loss of safety for profit continues to grow, which is scary.
Nothing I haven’t heard in some capacity before but managerialism as you can imagine is not good.
Nothing I haven’t heard in some capacity before but managerialism as you can imagine is not good.
Displaying 1 - 15 of 15 reviews