Conclusion

The dominant popular narrative about social change centers on individuals: Rosa Parks refused to give up her seat and the Civil Rights Movement was born. This narrative is factually wrong and strategically disabling. Rosa Parks was the secretary of the Montgomery chapter of the NAACP and had recently attended the Highlander Folk School, a training center for labor and civil rights organizers. The Montgomery Bus Boycott that followed her arrest was organized by the Montgomery Improvement Association, coordinated carpools across a city for over a year, and was sustained by the labor of hundreds of people whose names are not remembered. The reduction of that campaign to one person's spontaneous act of courage makes it both more inspiring and less useful as a model.

The most rigorous quantitative analysis of social change I know of is Erica Chenoweth and Maria Stephan's study of 323 resistance campaigns between 1900 and 2006 [Chenoweth2011]. They found that nonviolent campaigns succeeded roughly twice as often as violent ones, and that the threshold for success was consistent: campaigns that engaged the active participation of roughly 3.5% of the population did not fail. The mechanism is not mysterious. Nonviolent campaigns can recruit from a broader population, including people who will not take up arms but will march, boycott, strike, or withdraw labor. Broader participation creates broader legitimacy and makes it harder for the state to frame repression as protecting order rather than suppressing dissent. The 3.5% figure is not a guarantee; it describes a historical pattern. But it is a more useful starting point than the assumption that popular majorities produce change automatically.

Economic disruption is the mechanism that connects organized pressure to actual policy change. The Montgomery Bus Boycott worked because it destroyed the bus company's revenue from its Black ridership. The South African divestment campaign worked because it raised the cost of the apartheid state's international borrowing. In each case the mechanism was economic: the people in power faced a cost-benefit calculation that changed. Moral suasion may have affected some individuals. It did not change the structural calculation that drove policy.

Moral arguments have a poor track record as the primary lever of social change. They aren't irrelevant: they build coalitions, provide the framework that justifies what a movement is asking for, and affect potential participants' willingness to accept personal costs. But moral arguments addressed to those in power, without the economic or political pressure that makes their rejection costly, consistently fail. The tobacco industry did not voluntarily stop marketing cigarettes to children because public health advocates published articles about harm. Corporate privacy practices do not change because researchers demonstrate the extent of surveillance. What changes the behavior of those who benefit from a harmful arrangement is when not changing becomes more costly than changing.

Moral Reframing in Practice

The criminalization of homosexuality through sodomy laws is often treated as an ancient prejudice that modernity finally overcame, but its legal form is specific and recent. The British Offences Against the Person Act of 1861 and the Criminal Law Amendment Act of 1885 (under which Oscar Wilde was prosecuted) established the statutory framework that then spread through British colonial law to India, Kenya, Singapore, and dozens of other territories. This was a Victorian legal export, not a timeless human universal.

The Stonewall uprising in June 1969 catalyzed what followed, but that "what" was organized political work: building community institutions, litigating against discriminatory laws, lobbying legislatures, training activists, and making the case for rights in terms that demanded equal treatment rather than tolerance [Hirshman2012]. The AIDS crisis beginning in the early 1980s devastated these communities. ACT UP, founded in 1987, developed direct action tactics that accelerated drug approval timelines, targeting the FDA, pharmaceutical companies, and media coverage. In doing so, they changed what was considered acceptable activist pressure on public health institutions.

The crisis also produced an enormous wave of political organizing, legal advocacy, and coalition building driven by the urgency of survival. The LGBTQ+ movement that emerged from the AIDS years was more sophisticated, better organized, and more willing to engage in explicit political conflict than it had been [Faderman2015].

Legal change on LGBTQ+ rights did not arrive at once. Each step was criticized as premature by conservatives and inadequate by the communities affected, but each made the next one thinkable. Marriage equality, which seemed to most observers politically unreachable in 2000, was recognized by the US Supreme Court in 2015. The speed of that final sequence reflects both the sustained organizing of prior decades and specific features of the American constitutional system that allowed judicial strategy to outpace legislative consensus.

What these changes reveal about how moral consensus shifts is that the process has identifiable components. Changed facts matter: the AIDS epidemic made invisible communities visible in ways that shifted public sympathy and political calculation. Changed visibility matters: LGBTQ+ people in prominent public roles before and after coming out reduced the distance between the abstract policy debate and human recognition. Organized political pressure matters: neither marriage equality nor any other major rights advance arrived without sustained campaigns that built coalitions, litigated strategically, and maintained pressure across decades of partial progress and setback. Finally, legal strategy matters: advocates have to choose which cases to bring, in which courts, at which moment [Hannig2023, Hirshman2012].

The Expanding Circle

The circle of people we recognize as fully human has expanded over the last two and a half centuries—slowly, and with catastrophic reversals, but through mechanisms that are identifiable and replicable.

The historian Lynn Hunt points to something unexpected: novels [Hunt2007]. The eighteenth century saw an explosion of epistolary fiction (i.e., novels written as letters) that asked readers to inhabit the inner life of someone unlike themselves, to feel the world from inside another person's consciousness. Hunt argues that this trained readers to recognize the humanity of strangers and understand that other people's interior lives were as real as their own.

Whether novels were a cause or a symptom of changing attitudes is debatable, but slave narratives in the nineteenth century worked on the same principle. The Interesting Narrative of Olaudah Equiano (1789) and Frederick Douglass's Narrative (1845) put readers inside an enslaved person's experience in a way that statistics never could. The abolitionist movement distributed these books deliberately and widely. Making suffering something a reader from a comfortable life could imaginatively enter was a political strategy [Hochschild2006].

A second mechanism is organized political pressure. The Somerset case in Britain in 1772 illustrates how this works. James Somerset was an enslaved man brought to England who escaped and was recaptured by his enslaver. His case was taken up by abolitionists, who argued that English common law did not support slavery. Lord Mansfield's ruling was narrow—it said only that Somerset could not be forcibly transported out of England—but abolitionists used it as a wedge.

They had already built a network of committees, pamphlets, and petitions. The boycott of West Indian sugar in the 1790s was the first mass consumer boycott in British history. These campaigns eventually produced the Slave Trade Act of 1807 and the Slavery Abolition Act of 1833. The expansion of the moral circle, here, was not a spontaneous change of heart by elites. It was the result of sustained, organized work by people who built coalitions, applied pressure across decades, and framed their arguments in terms their opponents could not easily dismiss.

The expansion of the moral circle has never been linear. Reconstruction in the United States after the Civil War extended legal rights to formerly enslaved people, but was then systematically dismantled through paramilitary violence, electoral manipulation, and judicial capitulation. Most of the gains of the 1860s and 1870s were reversed by 1900. It took another sixty years, and another sustained organizing campaign, to recover them in formal law. The current rollback of rights for trans people in several countries is not an anomaly. Every extension of the moral circle has produced a counter-movement arguing that the extension is too far, too fast, or aimed at the wrong people. Sometimes those counter-movements win, at least for a while.

The relevance to technology is direct, even if the industry has largely avoided the conversation. Every system that makes decisions about people embeds assumptions about whose interests count and whose harm is legible. Facial recognition systems that perform poorly on darker skin tones are not neutral tools that happen to malfunction for some users—they encode a prior judgment about whose face needs to be recognized accurately. Content moderation systems that work better in English than in Amharic or Tagalog reflect a prior judgment about whose speech deserves moderation resources. Peter Singer argued that reason itself pushes in this direction: once you accept that suffering matters, it becomes harder to explain why only the suffering of people like you matters. That logical pressure is real, but it does not do this work on its own. Someone has to organize the campaign, draft the legislation, build the coalition, and be prepared to keep going when the counter-movement wins the next election [Singer1981].

Five Tools

The preceding essays have argued that large-scale harm is produced by systems with structural incentives to produce it, and those systems change when organized groups impose sustained economic or political costs on them. That argument implies that individual ethical consumption, career choices, and personal commitments matter to the people who make them, but are not the mechanism by which industries get regulated or power gets redistributed. Instead, five tools have historically accomplished this:

Collective organizing means building institutions with the capacity to act in coordination. Unions, professional associations, tenants' organizations, and consumer groups all operate on the same principle: the leverage available to any individual is small, but the leverage available to an organized group is proportional to how many members it has and how much they are prepared to do. In the tech industry, this includes formal union organizing, which has made progress at some major tech companies in recent years, but also informal networks, internal advocacy groups, and walkouts. The 2023 Writers Guild of America strike demonstrated that workers can negotiate explicit contractual limits on how AI is used against them—limits that no individual writer could have secured alone.

Regulatory pressure means participating in the political processes through which rules get made. This includes litigation, which has broken up monopolies, forced safety standards, and established legal rights. It includes lobbying, which is what organized groups do when they communicate with legislators. It includes public comment, which is how regulatory agencies collect input from people outside government. It includes electoral organizing, because the people who make and enforce rules are selected through elections. None of these are glamorous, and most of them are slow. The tobacco case took forty years from the first clear science to a partial legal settlement. The environmental victories described in earlier chapters took decades of sustained pressure. The tech industry's political connections are currently a source of strength; the history of the yakuza suggests they can be made a source of vulnerability through sustained political decisions to make the cost of association with harmful practices prohibitive.

Alternative structures means building institutions that demonstrate other models are possible. Worker cooperatives, public banking, open-source software, public payment infrastructure like India's UPI— these are not just alternatives to existing institutions. They are existence proofs that the existing institutions are not the only possible arrangement. Mondragón has been running for nearly seventy years. The Viennese Gemeindebau has housed working-class families for a century. Credit unions serve roughly a billion people worldwide. The claim that shareholder capitalism is the only workable model for large-scale enterprise is empirically false, and demonstrating that falseness is itself a form of political action.

Strategic communication means engaging seriously with the psychology of belief change. Facts do not change minds in contested domains; stories told by trusted messengers do. This is not a counsel of manipulation. It is a recognition that if you want to change what people believe about AI regulation, platform power, or labor rights, you need to find people whose social worlds already include such conversations, meet them where they are, and make updating their beliefs consistent with rather than threatening to their identities. Public health campaigns that have worked—on smoking, on seatbelt use, on HIV— did not do it by publishing better statistics. They found language and messengers that made the change feel possible.

Sustained participation means showing up repeatedly over years. The Federalist Society built its judicial pipeline over forty years. The labor movement built the eight-hour day over forty years. The LGBTQ+ rights movement built toward marriage equality over forty years. This is the part of the agenda that is most difficult to communicate to people who are accustomed to fast feedback loops. No single action produces structural change. What produces structural change is organizational capacity built before a political opening arrives, and the willingness to use that capacity when the opening comes.

Where Do We Start?

At work, the most immediate choices are about what to build and who to build it for. These essays have described how products encode assumptions about whose harm is legible and whose is not, and about whose data is worth protecting and whose can be extracted, packaged, and sold. Asking those questions explicitly in design reviews and sprint retrospectives and architecture discussions is a step toward change. It is not sufficient on its own, but it is not nothing. After all, every major organizational harm documented in these essays was produced by organizations full of people who did not ask these questions.

If your company has a union or a worker advocacy group, joining it is the highest-leverage individual action available. If it does not, it's worth remembering that the tech workers who have organized in recent years did so the same way everyone else has: by finding other people who shared their concerns and committing to sustained collective action. (It's also worth remembering that Google's cafeteria workers organized while the engineers were still debating whether it was "appropriate".)

In community, you may already be doing something valuable, like showing up at a literacy program or helping at a food bank. Libraries, community environmental groups, and other civic organizations are where the capacity for political action gets built. The Montgomery Bus Boycott was organized by people who knew each other from church, and the abolitionist movement in Britain by people who knew each other from Quaker meetings. The specific institution matters less than the sustained, repeated contact that allows people to build trust and coordinate action [Beckerman2022].

Politically, the most useful question is not "which party should I vote for?" but "where can I apply organized pressure most effectively?" Electoral organizing matters, but so does showing up to public comment periods, participating in professional associations that engage with regulation, and joining organizations that are doing the sustained work of advocacy on issues that are more specific than a party platform. The 3.5% threshold cited earlier is not an abstraction: in a city of half a million people, it is seventeen thousand people doing something instead of just complaining.

A Closing Note

The tech industry isn't as new or as revolutionary as it pretends to be. It is an industry with the same structural pressures, the same political dynamics, and the same aversion to regulation as every industry before it.

As I said in the introduction, I wrote this because most people in tech don't know enough to have an immune response to Silicon Valley's self-serving bullshit. This isn't just a matter of taste: decisions with consequences for billions of people are being made by a small number of accountable billionaires.

But I am not finished being hopeful. Change is hard, but not impossible. Gay marriage is legal. The ozone hole is closing, and the world is building solar power faster than anyone imagined was possible.

We are not going to solve big problems alone. We are going to have to work with people we disagree with, accept partial victories, and keep going after setbacks. We are going to have to be patient and organized and strategic, which feels much less satisfying than righteous anger.

But we already know how to build things. We know how to think about systems and debug complex interactions. The political and social systems these essays have described are not more complex than the technical systems we work with. They are just harder to test.

So: start where you are, use what you have, help who you can.