To Save Everything, Click Here: The Folly of Technological Solutionism

By Evgeny Morozov

11 min read

Why does it matter? Because the 'smart' future being built for you is quietly making you dumber, less free, and harder to change.

Imagine a city where crime is obsolete, obesity is cured, and politicians can no longer lie. Every surface is intelligent, every impulse is optimized, every problem has been quietly solved before you could notice it existed. Sounds like progress. Now ask the harder question: in a world where the environment itself enforces good behavior, who is actually being good? Morozov's answer is unsettling — nobody. Goodness requires the possibility of its opposite, and a civilization that engineers away every wrong turn has also engineered away the only meaningful kind of right one. Silicon Valley's crusade to fix everything is not philanthropy but a power grab — one that trades the friction of genuine human life for the smooth sensation of a problem you never got to solve yourself.

The 2020 Vision Sounds Perfect. That's the Problem.

Your glasses catch you reaching for the madeleine and Siri flags the glucose spike. Sorry, Marcel.

This is roughly the future Evgeny Morozov sketches at the opening of his attack on Silicon Valley's ruling ideology. By around 2020, the vision goes, self-tracking devices eliminate obesity and insomnia; smart glasses record your entire life so memory loss becomes a technical glitch rather than a human condition; and environments — walls, sidewalks, login screens — learn to detect criminal intent before anyone acts on it, making prisons as obsolete as the crimes they once housed. The pitch practically sells itself. Less suffering, less corruption, less waste. Who could object?

Morozov does, and his objection is precise. The gleaming 2020 scenario depends on a single substitution: human judgment traded for algorithmic management. When smart glasses log your glucose data and flag your childhood treat as a health hazard, the choice about whether to eat it has been quietly outsourced. You're no longer someone who decides; you're a node receiving an optimized recommendation. The Proustian experience — the ambivalence, the small surrender, the very human act of eating something you probably shouldn't — gets reclassified as a bug to be patched. It's worth sitting with that for a moment: the memory, the pleasure, and the mild guilt are not incidental to the madeleine. They are the madeleine.

Morozov's argument is that the bug is the point. The friction, the fallibility, the room to err and then reckon with having erred — these aren't obstacles to a good life. They're the raw material of moral agency. A system that prevents every mistake doesn't produce wiser people; it produces people who've never had to be wise. Which is what makes the solutionist vision of 2020 worth examining closely: it doesn't ask what we lose when every problem gets a technical fix, and that question turns out to be the whole ballgame.

'Solutionism' Is the Word You Didn't Know You Needed

Here is a word you need: solutionism. Once you have it, you will see it everywhere.

The concept comes from architecture criticism, where it names an obsession with grand, elegant fixes to problems that are actually messy, contested, and not fully understood. Morozov borrows it to describe the operating ideology of Silicon Valley — the reflex that treats every human situation as a puzzle awaiting the right algorithm. The trouble isn't that proposed solutions fail to work. The trouble is subtler: solutionism warps the problem before it solves it, so that by the time the solution lands, the original human situation has been quietly replaced by something flatter and more manageable.

Consider BinCam, a project from British and German researchers that makes a reasonable ambition — getting people to recycle — into a masterclass in what this warping looks like in practice. A smartphone mounted inside a trash lid photographs every item you throw away. Workers hired through Amazon's crowdsourcing platform categorize what they see. That data flows to Facebook, where your household earns gold bars and tree leaves based on how virtuously you dispose of your waste, and your neighbors can watch your score in real time.

Recycling, as it existed before BinCam, was a civic act — something you did because you believed you had obligations to people you'd never meet and a planet you shared. BinCam replaces that moral structure with a points system and a surveillance feed. Which is fine, right up until your Facebook friends stop watching, or the game gets boring, or the algorithm changes. The behavior that was supposed to be cultivated has no roots; it was propped up by incentives the whole time. Morozov's term for this is local optimization producing global damage: you solve the recycling problem in one neighborhood on one Tuesday, and you corrode the broader habit of acting well when no one is keeping score.

The deeper problem is that the researchers never ask whether recycling is the kind of problem that should be solved this way. They assume it is a technical coordination failure — people just need better feedback loops — and skip right to the mechanism. Solutionism reaches for the answer before the questions have been fully asked. What looks like problem-solving is actually problem-suppression: the messy human reality gets filed away, and a cleaner, gamified substitute takes its place. Civic trust doesn't survive that substitution. Once good behavior becomes a performance for an audience — a score, a leaderboard, a neighbor's approval — it has no purchase in a world where the audience disappears. The inefficiency, the ambiguity, the absence of any external reward: those aren't bugs. They are what makes the act count.

'The Internet' Doesn't Exist — and Pretending It Does Keeps You Powerless

Solutionism needs a delivery vehicle, and Internet-centrism is it. What if the phrase 'the Internet' — the thing everyone from senators to startup founders treats as a fact of nature — is actually a political move masquerading as a description? That's Morozov's central provocation, and once you take it seriously, a remarkable amount of confusion evaporates.

He calls the syndrome 'Internet-centrism': the habit of treating 'the Internet' as a unified entity with an inherent nature, a logic, a teleology — something that speaks to us if we'd only listen. Google's Eric Schmidt says policymakers should 'work with the grain of the Internet.' Harvard's Lawrence Lessig, confronting the erosion of privacy by transparency-gone-wrong, concludes that 'the network is not going away' and therefore we must accept its architecture and find legal workarounds instead. The network, in this framing, is gravity.

But watch what Lessig's move actually does. When California's Proposition 8 donors had their home addresses plotted on a public map — because a campaign-finance transparency law designed for dusty town-hall archives suddenly met a free Google Maps API — the resulting harassment tool wasn't gravity. It was a series of specific, reversible design choices: what data the API would serve, how the transparency law defined 'public,' whether disclosure records could be set to read-only or given expiration dates. Every one of those is a dial someone could turn. Lessig's response is to leave the dials alone and push for publicly funded elections instead, because tinkering with the infrastructure would violate 'the architecture of the Internet' — meaning, in plain English, that the design decisions baked into the network by specific companies and engineers are now off-limits for democratic revision. The network is sacred; everything else must bend.

What Internet-centrism costs you is the entire toolkit of targeted fixes, declared off-limits by metaphor. The same logic that protects a teenager's right to remix cat videos also forbids a municipality in Argentina from adding a CAPTCHA to its spending database to slow automated scraping. The same 'architecture' governing peer-to-peer file sharing supposedly governs whether court records should be indexed by search engines. Everything with an internet angle gets absorbed into one undifferentiated holy object, and any intervention — however specific, however reversible — gets dismissed as heresy.

Once you see that the plumbing is where all the interesting choices live, 'the Internet' stops being a force you must obey and becomes a collection of specific platforms, each shaped by specific people, for specific reasons — every one of which is, in principle, negotiable.

Democracy Needs Hypocrisy, Backroom Deals, and Other 'Bugs'

Democracy needs its vices. That's not a cynical concession — it's the operating principle that digital reformers keep destroying without realizing it.

Here's the test case. Germany's Pirate Party built a software platform called LiquidFeedback to let ordinary members propose policy and delegate votes to whoever they considered more expert on a given subject. The idea sounds genuinely appealing: finally, a mechanism that drags party decision-making out of smoke-filled rooms and into the open, where everyone participates and expertise finds its natural level. In practice, in a German federal state of nearly eighteen million people, twenty votes were cast on a contested circumcision law. Twenty. The platform didn't produce grassroots democracy; it produced a very sophisticated-looking suggestions box that almost nobody used.

But the deeper problem isn't the low turnout — it's the theory underneath. LiquidFeedback assumes that politics is a problem of information aggregation: build the right channel, and good preferences will flow through it. What this misses is that the hard part of politics has never been collecting opinions. It's bargaining between people with genuinely incompatible interests until something both sides can live with emerges from the wreckage. That process requires vague language that lets rivals claim they won, and the breathing space to reach a deal before every partial concession becomes a scandal.

A Polish dissident once described democracy as 'eternal imperfection, a mixture of sinfulness, saintliness, and monkey business.' He meant it as praise. The imperfection isn't a bug awaiting a patch; it's proof that real competing interests are being held in tension rather than steamrolled. When transparency reforms, fact-checking algorithms, and LiquidFeedback-style platforms all converge on making every compromise visible and every inconsistency scandalous, they don't clean up democracy. They make the actual work of democracy — the deals, the fudges, the deliberate vagueness — impossible to perform. What replaces it is theater: politicians who say nothing committable, or platforms that generate the appearance of participation while real decisions migrate somewhere less monitored. This is where Rosa Parks becomes relevant: the most consequential democratic acts tend to happen not through official channels but precisely because those channels are bypassed entirely.

Rosa Parks Would Never Have Reached the Front of the Bus

It is 1955. Rosa Parks boards a city bus in Montgomery, Alabama, and sits near the front. When the driver orders her to move for a white passenger, she refuses. She is arrested. The arrest becomes a lawsuit, the lawsuit becomes a movement, and the movement ends legal segregation on public transit across the country. The whole chain depends on one thing: the system was inefficient enough to let it happen.

Now run the scenario forward. Imagine a smarter bus. Sensors at the stop count passengers by race. An algorithm, drawing on historical data about which demographic combinations have produced 'tension incidents,' calculates safe boarding ratios in real time. A facial-recognition camera cross-references the social media profiles of people waiting in line and flags those statistically likely to cause disruption. Parks gets a polite text message: there's no room on this service, but another bus will arrive shortly. No confrontation. No arrest. No lawsuit. No movement. The system worked perfectly.

Morozov's point is that this represents a move from enforcing rules by appealing to people's moral sense or self-interest, to making certain behavior structurally impossible. You can argue that a law is right, or that it serves your interest — both approaches leave you the option to disagree. Or you can engineer away the option entirely. The first two leave friction. You might resist, suffer consequences, and in suffering them force a public reckoning with the law itself. The third erases that possibility. There is no arrest for Parks to leverage because there is no confrontation for the system to permit.

Bruce Schneier calls lawbreaking a catalyst for social change — the mechanism by which societies learn that a law no longer reflects what they actually believe. Virginia's anti-miscegenation statute fell because Mildred and Richard Loving violated it and were convicted. Texas's flag-burning prohibition was struck down because Gregory Lee Johnson burned a flag and the case reached the Supreme Court. You cannot challenge a law you are structurally prevented from breaking. The algorithm doesn't testify, doesn't get cross-examined, doesn't generate a headline.

The solutionist promise is that more efficiency means less injustice. What Morozov demonstrates is that efficiency and justice are not the same project — and that when you build a world frictionless enough to prevent every wrong, you also build one stable enough to preserve every wrong that has already been wired in.

The Personal Prospectus: How Sharing Your FitBit Data Traps Everyone Else

Legal scholar Scott Peppet calls it the personal prospectus: the idea that individuals should be able to package their behavioral data — FitBit readings, driving records, sleep patterns — and trade it directly with insurers and employers. For someone healthy, financially stable, and in possession of favorable data, this sounds like transparency finally working in the individual's favor.

But tilt the lens toward the person who is none of those things. Someone managing a chronic illness, working irregular hours, eating cheaply — their personal prospectus doesn't unlock discounts. It documents liability. And here is where the trap closes: once healthy, productive people discover they can trade their biometrics for lower premiums, a quiet norm crystallizes. People who share have nothing to hide; people who don't are presumably hiding something. Non-disclosure stops being a neutral choice and becomes an admission. Nobody passed a law requiring you to submit your health metrics. The coercion is structural, not legal — and it runs entirely on the voluntary decisions of people who benefit from sharing. Consider the neighbor who works two jobs and eats what she can afford and declines to share her data. She doesn't distrust the technology. She just knows what her numbers would say. In the new system, her silence is its own signal.

The collective coercion argument is what matters here, not the observation that self-tracking has downsides. Your decision to publish your health metrics changes the meaning of my decision not to. Privacy stops being a legal protection and becomes something you purchase if you can afford it and if your data is good enough. For everyone else, the personal prospectus isn't a tool of empowerment. It's the mechanism by which existing inequalities get laundered into actuarial science.

Smart Parking Meters Are Stealing Your Moral Education

You pull out of a Santa Monica parking spot with eight minutes still on the meter. Under the city's new sensor system, the meter detects your departure and resets immediately — your unused time evaporates, the city captures the revenue, and the driver pulling in behind you starts from zero. Efficient. Clean. And, Morozov argues, a quiet robbery.

Not of money. Of the chance to be a citizen.

Here's what the system replaced: before the sensors, a driver who left early might choose to let the clock run for whoever came next. Maybe she was feeling generous; maybe she just didn't want the city to have the money. Either way, she had to decide. The Santa Monica meter erases that decision by making it for her, and in doing so optimizes away the moral question entirely. Nobody is more virtuous as a result. The opportunity to be virtuous simply no longer exists.

Morozov's counterproposal is the real argument: imagine a meter that, when your departure is detected, asks you a question. Keep your remaining time running for the next driver — perhaps a grad student, the data might suggest — or reset and boost the city budget? Suddenly you're thinking about parking policy, about urban poverty, about who the infrastructure actually serves. The meter becomes an education rather than a vending machine.

This is the distinction Morozov builds everything toward: technology that maximizes economic efficiency versus technology designed to maximize what he calls deliberative efficiency — your capacity to think about the systems you inhabit. The difference isn't just philosophical. A citizen who's been forced to confront who subsidizes whom in a parking system tends to show up differently at a city council meeting about transit funding.

That habit of confrontation doesn't come from frictionless systems. It comes from ones that refuse to let you off the hook. Morozov's provocation is that the smartest possible system might be one that refuses to — that hands the question back to you, uncomfortably, and waits.

The Question Worth Carrying Out of Every App Store

Here is the real question Morozov leaves you with, and it's not about technology at all: who gets to answer the question of what a good life looks like — you, or the system built to optimize you? Every frictionless interface, every algorithm that reroutes the bus before the protest can happen, every meter that resets before you can decide to be generous — each one casts a quiet vote on your behalf. The adversarial alternatives he describes aren't nostalgic or anti-progress. They're a design philosophy that treats your discomfort as the product rather than the problem. The standard is simple enough to carry with you: when a technology makes you think harder about what you actually value, it's working for you. When it makes that question disappear, it's working instead of you.