Optimal Illusions: The False Promise of Optimization

by Coco Krumme

How optimization took over the world and the urgent case for a new approach

Optimization is the driving principle of our modern world. We now can manufacture, transport, and organize things more cheaply and faster than ever. Optimized models underlie everything from airline schedules to dating site matches. We strive for efficiency in our daily lives, obsessed with productivity and optimal performance. How did a mathematical concept take on such outsize cultural shape? And what is lost when efficiency is gained?

Optimal Illusions traces the fascinating history of optimization from its roots in America’s founding principles to its modern manifestations, found in colorful stories of oil tycoons, wildlife ecologists, Silicon Valley technologists, lifestyle gurus, sugar beet farmers, and poker players. Optimization is now deeply embedded in the technologies and assumptions that have come to comprise not only our material reality but what we make of it.

Coco Krumme’s work in mathematical modeling has made her acutely aware of optimization’s overreach. Streamlined systems are less resilient and more at risk of failure. They limit our options and narrow our perspectives. The malaise of living in an optimized society can feel profoundly inhumane. Optimal Illusions exposes the sizable bargains we have made in the name of optimization and asks us to consider what comes next.

Genres: Nonfiction, Science, Psychology, Technology, Philosophy, Audiobook, Business

256 pages, Hardcover

Published September 12, 2023

Reviews

[Miranda · Rating 3 out of 5]

⭐️⭐️⭐️I was very excited when I got this ARC as the limits and failures of Mathematical optimization is a topic I have a lot of interest in, and the idea behind this book has a lot of promise. Unfortunately, I found the book lacked something in the execution. It felt rushed and scatterbrained, like maybe it was supposed to be a collection of essays or case studies, but it got dropped on the floor and all jumbled up together. It hopped between memoir, interviewer notes, and historical study, never staying long enough to leave an impression. This book is packed with interesting tidbits and nifty facts, but needed a stronger editorial direction to really shine. It’s a crying shame, because the ideas and message are really important. We have a saying in mathematics, that one rarely finds the clearest explanation of a concept by the person who first discovers it. I think we’ll have to hope that there will be others who follow this path and tell a clearer story.

I received an Advance Review copy in exchange for this honest review.

#bookstagram #bookreview #backtoschool

[A Ter · Rating 3 out of 5]

First chapter has a few high points. Rest is unoriginal -- at times annoyingly so. This didn't need to be a book, there's really not enough for a book -- it would have been more compelling as a New Yorker article.

[Jason Braatz · Rating 1 out of 5]

This is a book that perfectly demonstrates the age-old wisdom of "don't judge a book by its cover." The anticipation of diving into this topic had me nearly shedding tears of joy and dancing the non-fiction bookworm dance (well, almost). I wanted an interesting analysis of various optimized "things" and how they either did or didn't live up to their explanations. Looking at false optima would have been a good start. Instead, I find that this is the author's polemic about American Bison, GMO foods, northern railroad history and dissatisfactions with cruise ships.

Before reading it, I was really rooting for this book. Honestly, I was prepared to start a fan club for it—if only it delivered. A total makeover by a new author on the same subject? Sign me up yesterday! The world is ripe for a sharp critique on optimization, a dive into its successes and failures across various landscapes.

Take Six Sigma & Edward Deming's revolutionary contributions to optimizing manufacturing, for instance. This topic deserves its own saga, yet it barely gets a nod. It's like missing the main act at a concert. And the other optimization extravaganzas? Completely overlooked. You'd think a mention of team and task optimization strategies like SCRUM, Agile, or Kanban would be too much to ask for. Spoiler: it was.

Out of nowhere in the middle of the book, the author takes us on a leap to the subject of Universal Basic Income—a hot topic, sure, but its connection to optimization is about as thin as Aunt Jane's lasagna recipe has to quantum computing. And there are so many other baffling detours—like a car ferry trip to the San Juan islands mixed with a puzzling disdain for cruises not running at full throttle. (This bit personally irked me, betraying a misunderstanding of basic fuel optimization in large maritime vessels that even a casual enthusiast could debunk).

And a swipe at Marie Kondō ; why, I ask? Your guess is as good as mine. A misplaced critique on optimization, perhaps? I can only assume the book was trying to declutter its own narrative by throwing out coherence.

Echoing the sentiments of Publisher's Weekly, the book feels "muddled." It's a shame, really, considering Krumme's background. You'd expect a mathematician and scientific consultant to navigate the concept of optimization like a pro. Yet, here we are.

The book has, at best, gone through a disjointed writing process, pieced together over years with bits from a never-realized short film that the author mentions early on. Despite being a 2023 release, many parts of it read like a time capsule from a bygone era, missing the mark on current events and even flubbing a story about an Amazon mega-hub visit.

So, if you're debating between spending 4-5 hours reading "Optimal Illusions" or that same amount of time memorizing your breakfast cereal ingredients, I'd say go for the cereal. At least you're guaranteed to walk away with something useful—like knowing your daily intake of riboflavin.

[Lloyd Earickson · Rating 2 out of 5]

As the saying goes, “when you have a hammer, everything’s a nail.”  In Krumme’s Optimal Illusions, we contemplate what happens when the hammer decides it is no longer content being a hammer, and would prefer to be a rock, instead.  This might be stretching the metaphor, but it captures the essence of Krumme’s perspective.  Once a data scientist working in optimization, she became disenchanted with optimization and retreated to a homestead, from which she works as a consultant, engages in, to use another phrase, “retrospective Russeauistic utopianism,” and writes confused and over-described books that aren’t really about optimization.

Optimal Illusions found a place on my reading list, and got me to read it relatively soon after I placed it there, because it is billed as a book in which a leading data scientist explores the flaws involved with optimization.  These are not false claims, per se, but they are somewhat misleading.  Don’t go into the book expecting math-forward investigations of optimization methods, or even a rigorous exploration of optimization itself.  Optimal Illusions is written by a data scientist, and it is about the flaws involved with optimization, but Krumme seems to mistake correlation for causation, attributing to optimization real modern malaises alongside subjective discontents.

Mathematically, optimization is all about finding maxima or minima.  If you can define the variables of a system, and ascribe functions to each of them, you can create a parametric, multivariate function in n-dimensions, and use a variety of techniques to find the peaks and valleys of the resulting topography.  Most of us do this routinely, and without the explicit mathematical formulation, with a handful of variables in our daily lives.  We might optimize the route we take for time spent, or the activities in which we engage for personal benefit.

That is the process of optimization, and my hope was that the book would discuss issues with the mathematical process, or with the ways we implement it.  Mathematical functions, especially the more complex and esoteric they become, and produce results that do not align with what we want from them, and I hoped that the book would parse, for instance, how advanced optimization formulae sometimes produce peculiar results.  Instead, Krumme’s ills, which she blames on optimization, are shortcomings of the people doing the optimizing.

Krumme cites recent supply chain issues as an example of optimization gone wrong, and in one sense that’s true.  The supply chains broke down because they were over-optimized in one or two dimensions at the expense of other dimensions, like resiliency.  However, that’s not really the fault of the optimization process, but of the choices made by the people doing the optimizing.  The optimization did exactly what it was supposed to, but its builders did not account for all of the variables or did not sufficiently weight some of them.

Think of it like going for a run.  Say your goal is to complete the run as fast as possible.  If the only factor you consider in your optimization is your running pace, then your optimization will tell you to sprint.  Sprinting will cover the distance the fastest, but if you’re trying to run a marathon, you’re going to have some issues.  Those issues aren’t the fault of the optimization, but of not considering enough variables when you built it.  If you factor in distance and what paces you can sustain and for how long, the optimization will give you a better picture of what your best average pace should be to finish the marathon in the fastest possible time.  All of which presupposes that the goal is to complete the run as fast as possible, but maybe you have a different goal.

Instead, maybe distance and speed don’t matter to you – you want to see the most interesting things when you go for your run.  You could assign weights to things of interest in a certain radius from your starting point, plug in your running characteristics, and have your optimizer provide you with a best route to run to see the most interesting things, perhaps with constraints like needing to finish it in a certain time or only run a certain distance.  Alternatively, you could have your optimizer optimize for distance, pace, and spots of interest, and the system would balance those objectives to provide a solution that is not the shortest, not the fastest, and not the most interesting, but has the maximum of all these factors.

Going back to visualizing optimizations like topography, imagine those three variables are functions plotted along three dimensions.  One dimension is speed (x), one dimension is distance (y), and one dimension is interest (z).  Each of these variables can be characterized as a mathematical function, and they will have some relationship.  If you want to find the solution that will offer you the fastest speed, the shortest distance, and the most interest, you would search for where x, y, and z have a shared maximum (or minimum, depending on how you structured your parameterizations).  That maximum will not be the highest x ever gets, or y ever gets, or z ever gets, but it is where all three of them are highest at the same time.

Optimization is a methodology, not a philosophy, but Krumme treats it like a philosophy and a way of life, and problematic at that.  She asserts that our entire framework of optimization is undermining our potential for human flourishing, harming our environment, and exacerbating societal ills.  These, though, are failures of specific optimizations, not of optimization as a process.  If you’re dissatisfied with your life, it is not the fault of optimization as a method, but of optimizing for the wrong things, consciously or unconsciously.

To Krumme’s credit, she acknowledges that neither leaning further into optimization nor attempting to return to some earlier, notional, Edenic state will address the discontent she feels or the malaises she sees, but her proposed solutions are incoherent at best because, as already emphasized, optimization is not the cause of these states.  Optimization is a tool, not an organizational principle or a philosophy; using it to fill those roles is akin to using sandpaper to cut down a redwood.

Thus, much of Optimal Illusions feels like it’s dancing around the main issue without ever naming it, or even quite identifying it.  Its attempts at personal anecdotes and flowery language feel forced and unnatural, and the overall product is, well, suboptimal.  Though it highlights numerous, surface-level conditions, it does not follow up on those ideas.  The end result is a book that reads like the author listing perceived problems without understanding the underlying ideas or even why they are perceived as problems, much less suggesting solutions.  Optimization might be Krumme’s hammer, but for her readers, the world isn’t made of nails.

[Christopher · Rating 2 out of 5]

Really a 2.5.
I so so so wanted to like this book, and I feel like it has a finger on something pulsing through our society, but my god it was a mess. Nice insightful anecdotes and nuggets were lost in a narrative that whiplashed between an impressive collection of stories that didn’t all quite hang together. It’s relatively short, so may be worth the risk to scratch a curious itch, but you’re not missing much if you leave this one on the shelf.

[Blake Sooter · Rating 5 out of 5]

Wow. Really good. A careful consideration of the world of “optimal everything” that we are creating and the good and bad that comes with it.

[Jack Gihl · Rating 4 out of 5]

Really good ideas but at times I felt lacked cohesion.

[Geoff Thomas · Rating 4 out of 5]

Important analysis of our American penchant to seek optimization in every facet of our lives and society. Both positive and negative impacts are discussed at length with the author's' personal experience blended into her story.

[Canyen Heimuli · Rating 3 out of 5]

I feel deeply that there was (is?) potential for a really phenomenal book in this book’s core idea. But by lacking some clear editorial direction, it left me confused and unimpressed. The book could have been an article, or maybe it should have been 3x longer. I would have loved either a lot more than what I read.

By using the very specific term “optimization” in the title, the author, a former data scientist with an MIT education, purports to lay out a brief history of this specific subfield of mathematics dealing with dynamically deriving maxima or minima, and how it came to tentacle its way into so much of today’s predictive analytics and even social consciousness. The author kicks off the book by loosely defining “optimization” using an example that specifically refers to linear programming, but more broadly as the analysis of “what’s the best possible solution given some constraints we have?” Though I am not a layperson, I found the author did do a passable job at making this kind of analysis “sexy” and worthwhile to people who aren’t math nerds, at least for the length of an airplane trip. It’s cool to learn that if you type in, for example, a bunch of information about some foods you have access to and their quantities, Microsoft Excel, using basic math, can write you a diet that maximizes nutrition while still having some variety built in.

And where this book shines the most is both how it deconstructs what’s necessary for optimization as a formal analytical approach, and critiquing how society uses the language of optimization to approach things we intuit, or learn by sad experience, shouldn’t be optimized. I loved the author’s breaking down the concepts of atomization, abstraction, and automation; and I was taken by her descriptions of the tragedy of us trying to optimize dynamic, living processes like dating, going on a vacation, or saving the buffalo.

The main place where the book never really takes off into the territory of phenomenal work is in how it’s so general as to be very arduous to mentally construct a critical throughline. Perhaps it’s because the same generalist way she talks about “optimization”, and not in terms of the very specific subfield of math that it is, to make it more accessible is the thing keeping her from saying anything specific and definitive. Most chapters start with a description of some process or system, some eminently gone awry but most working as intended, and says to kick off the chapter some version of the phrase “this is the result of optimization in practice.” The author says this to describe all kinds of real world phenomena and systems, from farming, to how casinos are built, to how global supply chains run, and even to how Marie Kondo got so damned popular and bought a house in LA. I wondered aloud going from chapter to chapter “Is she still referring to optimization as trying to maximize something, or just the act of tracking something? Is just monitoring metrics optimization? Is any use of math, or an abstract model optimization? Is anyone that wants things to be better and trying to make things better essentially optimizing? And how can we de-platform them???” In this sense, at least, the book feels artificially contrived in how it often pretends away the distinction between analytically solving some business or technical problem using optimization, and using a similar philosophy to try to “optimize” more chaotic processes like life, love, joy, the universe, and what color of Mitsubishi Eclipse you should buy. The irony is that most optimization algorithms are sufficiently advanced that one actually can maximize, say, profits or leftover capital SUBJECT TO being sustainable, having as low-as-possible turnover, social responsibility, etc.

This relates to one other huge thing the author overlooks, or doesn’t consider with enough scope: the argument for WHY western thought so wholeheartedly adopted the “philosophy of optimization” (by which the author also apparently means “growth” or “efficiency at all costs”) is really narrow — both in time and in space. The author states in Chapter 2 that she bookends her historical analysis of the past and past ideas in optimization to Isaac Newton and his contributions to atomization and abstraction in particular. If this book were focused more on math, this would make a lot of sense as calculus underpins optimization. But, maybe unintentionally, this scope permits a number of western philosophies (such as capitalism, and liberalism) to be seen as foregone or taken as givens in the book’s scope. Most importantly, in the scope of this book, it doesn’t matter where capitalism came from: it was firmly established by the time of Newton so it’s taken as a given in the purview of this book.

This is illustrated, among other places, by the really bizarre (paraphrased) line in the book that “Whereas capitalism starts with parameters and lets the market work out a solution, optimization starts with a solution and finds values for the parameters. Capitalism is bottom up while optimization is top down.” (it should be noted that in the same paragraph the author uses the terms “market(s)”, “capitalism”, and “market capitalism” seemingly interchangeably) Not only does nothing else in the book accord with the idea that optimization and capitalism are diametric opposites, but the books doesn’t explore that capitalism is demonstrably one of the worst offenders of all the bad kinds of optimizations the author said she would write about. Long before Milton Friedman and “Trickle Down Economics”, capitalism has been defined as a profit and wealth maximization political philosophy that doesn’t just coincide with but depends on growth at all costs. This is not a Marxist view: it is the view capitalists themselves are happy to confess. Capitalism specifically, not just optimization broadly, is what determined to create enclosures, to enact chattel slavery, to colonize foreign countries, and now enforces debt bondage, among other things, and it absolutely used optimization to justify and more efficiently carry out these things. There is no other system besides capitalism that has so wholeheartedly, and to the entire planet’s detriment, forced the language and tenets of optimization into nearly every system of the modern world. Why the author doesn’t note this (at one point she writes “optimization wasn’t supposed to create a big winners and even bigger losers”) is extremely confusing and leads to awkward questions about whether the author is ignorant of this past or intentionally excludes it from the scope of her book and why.

One could (and maybe should) just as well examine the similarities between the language of optimization and that of parochialism, patriarchy, racism and social darwinism, and eugenics, to name a few. I feel as though if the author had done this, there would have been more material with which to answer the question of “how do we de-optimize?” and “How should we instead decide on things?”

There was once another timeline back before I knew what this book contained in which I read a book that did get into the nitty gritty of optimization which laid the foundation for the author’s philosophical arguments. That book might have discussed the differences between convex and non-convex optimization, vertices of polytopes, and why those weird little details matter in modern data science and modern society. Because the book I did read had some good information, and because I still have hope in what this book could have been and maybe what this author could write in the future, I give this book 3 stars.

I’m curious to see what she writes next!

[Tracie Hall · Rating 3 out of 5]

Optimal Illusions by Coco Krumme

BIBLIOGRAPHIC DETAILS
-PRINT: © September 12, 2023; ‎978-0593331118; Riverhead Books; 256 pages; unabridged (Hardcover Info from Amazon.com)
-DIGITAL: © September 12, 2023; Riverhead Books; 978059333132; 250 pages; unabridged (Digital version info from Amazon.com)
- *AUDIO: © 11 September 2023; Books on Tape; 7 hours (approx..); unabridged (Audio info from Libby app.)

-FILM: No

SERIES: No.

CHARACTERS:
N/A

SUMMARY/ EVALUATION:
-SELECTED: Perhaps I was searching for an economics book when this popped up and I decided to check it out and download. It wasn’t very long ago, but I have no memory of how this arrived on my virtual Libby shelf.
-ABOUT: The author describes a world built on the recommendations of computer models which assist in developing economic efficiencies toward an end of optimizing resources in the production of goods and services.
-OVERALL IMPRESSION: Coco has an impressive mathematical and scientific background, and a great vocabulary.
I get the sense that she herself developed (maybe still does) models for efficiency and optimal gain, so I don’t think she is anti-optimization (a concept I wasn’t familiar with until reading this), but perhaps a bit disillusioned, and certainly intent on explaining why we should consider the possibility that just because processes and procedures can be economically rewarding, doesn’t always mean the end result will be balanced or wholesome, in fact, I think she’s convinced that it OFTEN isn’t.
I enjoyed learning of her interviews with individuals as she asked about their experiences within their industries. Like the reluctant farmer, Bob, resisting the temptations of more lucrative practices that felt like betrayals to his values, until he eventually got swept up with the tide.
I felt I learned from her perspective of Las Vegas-style entertaining, railroad constructing, and more.
If I were to apply symbolism to the mood the book instilled, it would be the image on the Rider-Waite Tarot version of the 5 of cups where a wizened gray-haired figure cloaked in black stands gazing at three prone cups, their contents spilled; while two upright cups stand behind the figure, which I would say in this case would represent a nostalgic preference to ignore what’s been gained from the technology and science of optimization (the two standing cups) and focus on those slivers of humanity, spirituality, community spirit, and the like, that feel lost to the sands of time.
I came away mentally humming, “They Paved Paradise and Put Up a Parking Lot.”

AUTHOR:
Coco Krumme. From cocofolio-dot-com-slash-bio:
“I'm an applied mathematician and writer. My book OPTIMAL ILLUSIONS comes out in September.

I have a phd from MIT and spent time in academia and tech before starting Leeward Co, a consultancy that works with scientific r&d teams.

These days I live on a rural island where I run a craft distillery, make sculptures, and am learning to fly a small airplane. Occasionally I write about other topics, like why Silicon valley has no sense of humor. If you're looking for the S.V. Poetry Magnets, you can find those here. To get in touch about speaking or consulting, try leeward at cocofolio dot com.”

NARRATOR: Coco Krumme (see above)

GENRE: Non-fiction; economics; Technology; Psychology; Philosophy; Science; Business

LOCATIONS: (not all-inclusive)
Kentucky; Washington; New York; California

TIME FRAME: Contemporary and historical

SUBJECTS: (not all-inclusive)
Industry; Construction; Farming; GMO; Progress; Economics; Values; Buffalo; Optimization; Technology; Mathematical models

DEDICATION:
Not found

SAMPLE QUOTATION: Excerpt From “Introduction”
“In the middle of the field, the bulldozer pauses, and out pops its driver for a cigarette break. It’s a perfect September day in northern Kentucky: crisp, blue, free. A neighboring dozer carves ant-hill ruts through the construction site. In the distance are horse pastures and strip malls.
I weave past the open gate up one of the bulldozed roads until I reach the end of an ant rut. I park my small pickup truck, windows down, and let the dog out for a walk. The sun warms my face and my thoughts are lost in the dirt canvas, until suddenly, I feel the driver’s eyes fix on me. This isn’t any anthill to explore. He crushes out a cigarette and settles his gaze again in my direction, indifferent now, before turning back toward his rig and the work ahead.
Some say America’s too new a country to have ruins. I say, our ruins are just better hidden, under the sheen of the new.
In Athens, the Parthenon stands tall and crumbling, like a drive-in movie screen strung high above the tourists and café’ dwellers below. In Rome, the Colosseum anchors a swirling city, its cold stone presence cutting through the chatter of tour guides and the grease of chap pizza and the plastic trinkets made overseas. At Teotihuacan, you can clamber up each imposing pyramid and practically touch the scorching sky.
In America, by contrast, we hide our ruins in the desert or cart them off to Chinese scrapyards. After September 11, the remnants of the World Trade Center were wrapped like widows in black canvas , quietly disassembled, and sent abroad. An engineer named Cao Xianggen says to the Chicago Tribune, of the injured steel: “America can’t use it all, but China has a huge demand.”
Now I watch the bulldozer driver carve new ruins into his field. He’s helping build the Amazon Air Hub, a $1.5-billion site expected to host some one hundred cargo airplanes, three hundred trucks, and a robotic sort center sprawling over a million square feet. Slated to open in 2021, it’s currently a mess of construction equipment and barricades.”

RATING:
3 stars. It's perfectly well written and interesting. But I rarely give 5 star ratings, so my 4 stars are usually fairly riveting reads. 3 stars isn't bad, it just means it's not a subject that I get enthusiastic about.

STARTED-FINISHED
12/11/2023-12/14/2023

[Abram · Rating 1 out of 5]

It's hard to read a book this bad about something that has overlap with a deeply held belief. I was almost overjoyed to find this book in a public setting where people can pick it up and read it and discover issues with mistaking the tool of optimization for a full ideology, but now I just hope it's tucked away somewhere to be forgotten.

In no particular order of grievance, lest we be led into optimization, the following objectives:

- For all its talk of global optimization, this book is almost exclusively about the American West. Anyone even passingly familiar with how the West was created will know that you can't look at a landscape that was formed mostly by federal intervention and draw conclusions even about the country, much less the world. The fact is Krumme is conflating geography and ideology, while managing to fail to operate consistently across scale.

- Many of the examples of "escaping" optimization can be reframed as optimizing for different things, ie changing the objective function. Want to work less and spend more time in your community? Congratulations, you've optimized, and the fact that you use different language to describe it does very little to change actual facts and behavior.

- The book is almost insanely time-bound in its outlook. I don't think I've ever heard anyone insinuate before that creating the American West was an easy endeavor, but once it's tossed under the catchall of optimization, I guess it is! This book can be taken as a great-great-great-grandchild of the Enlightenment standing atop the mound of cultural and economic dominance that has been built up over centuries and bemoaning the fact that mound-building is just so easy. Furthermore, there is the implication that prior to Newton no one ever optimized for anything in a way we'd recognize as "flawed" today, all the megafauna we hunted to extinction and the centuries-old degradation of rice paddies in East Asia not withstanding.

- Her criticisms (mostly reasonable) of modern industrial agriculture fall wildly short of acceptability because a billion people would have died if it weren't for Norman Borlaug and the Green Revolution. Whether or not we have managed that revolution well is an entirely different question, and the emphasis on Borlaug is just as bizarre and great-man-fallacy adjacent as the strange claim that Descartes, after providing various proofs of the existence of God, was removing the divine from his explanation of the world.

- Never has a book made me more aware of the benefits of religion in blunting despair. I'd advise the author to simply consistently attend a local worship service, regardless of belief, and see the change.

- Finally, this entire book feels like a bad mirror of the philosophy put forth in Antifragile, by Taleb. It's not a fair comparison, they aren't the same, they do not have the same mission or audience or scope, but if you'd like to be convinced of this and have something that is fun to read, go read that instead.

[atom_box Evan G]

The book got me thinking beyond its covers. Since Optimal Illusions invited us to consider adding some skepticism to our optimization, I thought about who follows optimization and who doesn't.

On a scale of 100, where some groups might fall in their use of optimization:

0 - Eastern religions (Life is suffering; choose your pain. Don't optimize your escape from suffering. Don't subscribe to wellness influencers.)
90 - Christianity/Islam (there's an all-or-nothing pass/fail final exam)
60 - Judaism (lack of strong heaven-centricity, so less to optimize around, foregrounding the capricity of the deity)
100 - Effective Altruism (Donate 1 kidney; brokers buying mosquito nets.)
60 - Quakers (God is still speaking - could speak now, in the middle of this meeting, so there is not a 2000 year old lock on the OKRs)
20 - Math disciplines like Chaos, non-linear fluid dynamics, Game Theory, Network Effects and Emergence
100+ - Calculus
92 - Jeff Bezos
50 - Jack Dorsey
80 - Trump: few sincere beliefs or original ideas; just spitballing, watching TV, and a nearly-A.I. level of devotion to A|B testing to reach his OKR's.
60 - The Grateful Dead: hippies, sure, but their live show is optimizing several things. They're not really doing "whatever" up there.
30 - World Music DJ Jonathan Oberby on WPR - Fun show! His playlist really does achieve a "whatever" quality. But ironically it misses its intended 0 mark because it's very evenly distributed across genres, tempos, geographies. Oberby optimizes around lack of bias. But this introduces bias: the first thing you learn in freshman statistics is that a truly random coin will have runs of 10 identical results in a row
30 Me
80 Extraction focused colonialism (Haiti, Nigeria, Dutch East India) - there is an element of shame and moralizing that act as a brake on the all-out extraction velocity
70 - Easter Island: more tikis!
100 - petroleum-based agriculture (Coco discusses)
80 - destroying the buffalo of the Great Plains ( if Y Combinator or efficient venture capital had existed back then, much faster).
80 short-sighted urban planning in Atlanta (Coco discusses)
60 - Coco Krumme's Optimal Illusion. Krumme's still a fan of inflection points -- just with caveats. There are optimal ways to be non-optimal.

I only saw the author discuss the book at a book festival, I did not read it. I attended but was in the mood for a refutation of our practice of chasing "the best", as in "10 ways to hack your sleep". But this was the mildest of retreats from optimization, just a cautionary note. And that is fine. Probably optimal.

[Preston Gardner · Rating 5 out of 5]

like stuff I was thinking about myself, only clearer

I was taught optimization in school and then I worked on compilers. That was 40 years ago. Later I worked at various times, for Apple, Amazon and 9 different startups. So in some ways I come from a similar background to Coco Brumme’s, but I am older.

She covers all the important ground in her book.

There are several ways to understand satisficing, the term Herbert Simon used. Brumme describes it as stuff you do to patch up a model when your optimizations have outcomes you don’t like. But I read some other author who said natural selection does not optimize, but satisfices, and the main rule to satisfy is “don’t die”. You don’t have to be the best at anything: just don’t die.

My own spin on her book is that we and future humans will live in the junkyards of today, and reuse and recycling from those junkyards will be the focus of our lives. That includes media junkyards like old libraries.

(And perhaps you should ignore the five stars and wait for cheap used copies of this book to come out in three or four years. Or get it from your public library. Who knows what else you might find, if you wandered through the stacks of your library? Or shopped at a Goodwill Store? Or scavenged stuff from a dump?)

[Lisa notes · Rating 3 out of 5]

I like to be efficient. But after reading this book, I wonder if I should back away a little from worshiping efficiency. Krumme makes an excellent case that it's possible to take efficiency too far.

Krumme writes, "By embracing efficiency above all, we’ve crowded out what can’t be measured and optimized and allowed the metaphor of optimization to cannibalize other worldviews."

For example, consider our reliance on GPS for traveling. As someone who wants to know the quickest route to where I'm going, I have lost my "sense of adventure, as well as [my] comfort with getting lost."

After reading this book, I won't stop using GPS; I still want to know the optimal way to get somewhere. But I will think twice about being so dependent upon it. There are things in life that optimization can't get us. I don't want to lose those things at the altar of optimal illusions.

I recommend this book but with some caveats. Krumme gets a little too detailed for my taste with many of her examples. I lost the thread at times and got bored. Her premise is good, but it didn't always stay centered.

Nonetheless, Optimal Illusions is an intriguing read and an even more intriguing principle to ponder.

My thanks to Netgalley and Riverhead Books for the review copy of this book.

[Sarah · Rating 3 out of 5]

I'm not entirely sure what I think after reading this book. It was certainly engaging and thought-provoking, and helped me to see how my desire and drive for optimization had many upsides, as well as many downsides. There were certain chapters, such as the reintroduction of bison in the American West, that had new-to-me topics and I loved learning about them. But as good as many of the examples were, the chapters hopped around from riding trains to a bakery on the San Juan Islands, and the through-lines must have been more obvious to the author than to me. I didn't expect a neat conclusion or a "do these things" list, but the book just didn't hang together as a whole, nor did the individual vignettes stand strong enough alone. I do think parts of it were thought-provoking, so I'll be generous and stick with the 3 stars.

[Matt Stevens · Rating 3 out of 5]

Audiobook. Meh. I wanted to agree with her case. But she goes into the book having assumed she doesn't have to work that hard to prove that "Optimization is done". And she never does.
I compare this to "Blood in the Machine" by Brian Merchant that I just read where he spends an incredible amount of time doing the research into the Luddites and really brilliantly makes the case they weren't anti technology. The luddites were against the capitalists using the tools of the state and technology to impoverish everyone else and make themselves excessively rich and powerful.
Krumme fails to make a similar case. This book feels like the conclusion to what should have been a much more thorough run through of business cases.

[Kevin · Rating 3 out of 5]

Intriguing topic, got me to pick up the book. I wanted to like this book. Difficult to read though because I found it repetitive, lacking in structure, and easily distracted by unnecessary prose. Anecdotes are helpful to understand the points but I found them scattered and needing to ask myself, why is the author going into such lengths about this story again? Another reviewer suggests this could have been a New Yorker article and I agree that would’ve been a more effective way to get the message across.

[Jeevan · Rating 5 out of 5]

I started this book a few months back and picked it up a few days back and couldn't it put it own until I finished it. Something about the way the author described the mathematical intuitions behind optimization and the good and bad of the results - kept me hooked. The intertwining of characters alive and historical was impressive. I could not resonate with the categorizations of slack, pace, etc and neither did the last chapter on what else we could be doing. Regardless, kudos to the research and the walk through of the journey.

[Matt Heavner · Rating 3 out of 5]

Interesting look and perspective on the over dependence or all consuming paradigm of optimization. This is a synthesis of lots of fun/good perspectives (Ulam, Grothus!, Isaac Newton, Kurt Vonnegut, Borges, and many more). This did discussion the failures of optimization and the challenge of finding another paradigm when we are consumed with optimism? How do we do better in using optimism? ;) Or what's better than optimism? The path to the solution lies is recognizing the paradox, but no solution is giving.

[Joseph R · Rating 2 out of 5]

This book intrigued me at first due to the possibility of learning more about the downsides of optimization. While the author detailed a few interesting interactions with people who have experienced some hardship that could potentially be connected to the idea of optimization, I believe the author did not adequately make the case that these downsides are attributable to optimization. The examples expressed in the book could more accurately be attributed to more influential factors than the idea of optimization in my opinion.

[Lawrence · Rating 2 out of 5]

The author seems to blame all the ills of the modern American economy on optimization, ignoring such obvious villains as greed and increased industrial concentration. The author also seems to be less aware of the amthematics of optimization than claimed, as at one point she discusses linear programming and one page and then mention the simplex algorithm on the next as if it were a separate item, when in fact simplex is the algorithm that lies at the heat of linear programming.

[Harshal Patil · Rating 1 out of 5]

I picked up this book because it promised to expose the flaws in capitalism's interest in optimizing the world and maximizing productivity. I thought it would help me reduce my desire for perfection. But it didn't make a compelling case for me. I understand we should have slack in the system to withstand shocks or black swans. We can make our systems anti-fragile. But, I did not feel compelled to change anything in my life.

[Morgan · Rating 4 out of 5]

*3.75
Optimal Illusions takes an interesting look at the uses (and failures) of optimization in multiple fields. I really enjoyed the history of where this idea came from as well as how embedded it is in technologies. Overall, it’s a great read but the ending left me wanting a bit.

[Catherine Sherling · Rating 2 out of 5]

i didn't feel like this book showed me a single meaningful point honestly. it felt hollow - every point that was made was a .....well duh. yeah. that's how the world is. like i finished it feeling like there was no impact womp womp

[Rachel Kayman · Rating 4 out of 5]

Such an anti Rachel book … basically all the reasons why optimization is bad. Super well written, decent amount of jargon broken down for more general audiences. Some valid points are made but no conclusion / alternative solution was reached…

[Armen Shirvanian]

My interview with Coco Krumme on "Optimal Illusions" for The Armen Show:

http://www.armenshirvanian.com/podcas...

[Ridge Knapp · Rating 3 out of 5]

repetitive & at times lacking depth. Gets better towards the end, honestly if I'd started reading at chapter 5 I would have /considered/ a 4/5.

[Nina · Rating 2 out of 5]

fun enough for the particular examples used, but not really put together into a compelling narrative with an unshakeable thesis

[Arthur Krause · Rating 3 out of 5]

The book is interesting. The examples are good. The Voice of the reader is nice. The Conclusion is good. And still somehow it didn’t catch me so much. Anyway, good learnings.

[Richard · Rating 2 out of 5]

So boring. This genre of book is always such a hard slog.