Language Matters Podcast
A person opens his phone and sees that everyone is angry. Everyone is saying the same thing. Everyone is mocking the same target. Everyone is repeating the same accusation, the same slogan, the same moral certainty. A politician has betrayed the country. A migrant has ruined the neighborhood. A woman has lied. A product has changed lives. A war must be fought. A man must be destroyed. A company must be trusted. A nation must be hated. The screen says this is society. But who is speaking? Is it a thousand real people, each arriving independently at the same judgment? Is it one political campaign with a thousand accounts? Is it a marketing agency? A foreign government? A bored teenager? A company protecting its stock price? A bot farm? A group of paid influencers? An artificial intelligence system producing human-sounding outrage at machine speed? Is it a crowd, or only the costume of a crowd? A false claim deceives the intellect. A fake crowd deceives the social instinct. This is the quiet terror of the modern public square. We no longer know when society is speaking and when society is being simulated. That uncertainty does something to the soul. It does not merely confuse us about facts. It weakens our ability to trust our own perception of reality. We look at a comment section and wonder whether it is human. We look at a trend and wonder whether it is purchased. We look at a viral outrage and wonder whether it began in conscience or in strategy. We look at a wave of reviews, likes, replies, shares, slogans, flags, denunciations, and praise, and we ask the question no healthy society should have to ask every hour of the day: Is this the voice of the people, or is this machinery wearing the people’s face? The old internet had a promise. It was not always noble, and it was never innocent, but it contained a strange democratic grace. You could arrive without your passport. You could speak under a name you chose. You could become a handle, an avatar, a sentence, a recurring tone in a forum, a mind without a résumé. You were not always your job, your family, your class, your country, your body, your legal record, your employer, your accent, your address, your face. You could wear a mask. And the mask was not always a lie. Sometimes the mask was what allowed the truth to appear. A worker criticizing an employer may need a mask. A dissident under a regime may need a mask. An immigrant afraid of both the country he left and the country he entered may need a mask. An abuse victim may need a mask. A teenager discovering forbidden thoughts may need a mask. A person writing about addiction, sex, shame, grief, faith, betrayal, or spiritual collapse may need a mask. The mask can be cowardice, yes. But it can also be mercy. It can be the narrow doorway through which an endangered truth enters the world. The powerful rarely understand this. They often confuse exposure with virtue. They say, “If you have nothing to hide, use your real name.” But this is the language of people who have been protected by names rather than hunted through them. A name does not mean the same thing for everyone. One person’s real name is a platform. Another person’s real name is a leash. The real-name internet has always presented itself as moral hygiene. If everyone used their legal identity, we are told, there would be less cruelty, less fraud, less abuse, less chaos. There is a partial truth there. Some people become uglier when they believe they cannot be found. Some lies breed in darkness. Some threats should not be protected by pseudonyms. But the real-name solution is morally crude. It solves one problem by creating a larger one. It says: because some people abuse masks, no one should have them. It treats the whistleblower and the troll as the same kind of creature. It treats the dissident and the scammer as if both are merely hiding. It forgets that the powerful already have institutions, lawyers, security teams, public relations departments, citizenship, wealth, and distance. It is the vulnerable who need obscurity. The answer to fake people cannot be forcing all real people to become visible. That is the first principle. The second is this: anonymity is ancient, but infinite artificial multiplicity is new. Human beings have always hidden. We have signed pamphlets under false names. We have written letters anonymously. We have whispered against kings, churches, fathers, bosses, mobs, parties, and police. The hidden voice belongs to political history, religious history, literary history, and the history of survival itself. But something has changed. The mask used to belong to a person. Now the mask can be mass-produced. A single actor can manufacture thousands of apparent speakers. A company can create artificial praise. A political movement can simulate grassroots anger. A government can seed panic into another country’s public square. A scam network can flood reviews. A botnet can make fringe sentiment appear mainstream. An artificial intelligence system can generate endless comments, replies, profiles, images, biographies, confessions, jokes, prayers, accusations, and testimonies. The fake account no longer needs to sound fake. The artificial voice no longer needs to stumble. The machine can produce warmth, indignation, irony, grief, patriotism, moral certainty, consumer enthusiasm, ideological purity, and personal anecdote. It can say “as a mother,” “as a veteran,” “as an immigrant,” “as a teacher,” “as someone who used to believe the opposite.” It can borrow every costume of human credibility. The deepest fake is not a fake image or a fake quote. The deepest fake is a fake public. A fake public is more dangerous than a false statement because it does not merely tell us what to believe. It tells us what others already believe. It manufactures social reality. It surrounds the individual with an illusion of consensus. It says: everyone knows this, everyone sees this, everyone agrees, everyone is angry, everyone is laughing, everyone has moved on. And because human beings are social animals, because we are built to sense the tribe, because moral courage is exhausting and loneliness hurts, counterfeit consensus can become a form of governance. People do not only obey laws. They obey atmospheres. If the atmosphere can be manufactured, then power no longer needs to persuade each person directly. It only needs to make each person feel alone. This is the actual crisis. It is not simply that bots exist. It is that we are losing the ability to distinguish between a people and a simulation of a people. The crisis of the next internet is not only false information. It is false social reality. Imagine a city waking up to panic. Overnight, thousands of posts appear claiming that migrants have made the streets unsafe. The stories sound local. They mention neighborhoods, schools, gas stations, grocery stores, police scanners, worried mothers, old men afraid to walk at night. Some accounts have profile pictures. Some have years of ordinary posts. Some tell little stories with human details: a daughter who no longer takes the bus, a grandmother who heard shouting, a neighbor who “finally said what everyone is thinking.” A few of the posts come from real frightened residents. Some come from political operatives. Some come from newly created accounts. Some are generated by AI. Some are copied and localized across cities. Some are paid. Some are automated. Some are human beings reacting sincerely to a panic that was manufactured before it reached them. By morning, the trend is visible. By afternoon, politicians cite “public concern.” By evening, local news reports “growing outrage.” By the end of the week, a policy is proposed. The crowd has become real in its consequences, even if it was partially fake in its origin. This is how artificial posts become perceived consensus. Perceived consensus becomes media coverage. Media coverage becomes political pressure. Political pressure becomes law. The machinery creates the atmosphere, and the atmosphere governs the human being. So what should be done? The easiest answer is the worst one: make everyone prove who they are. Upload your passport. Use your legal name. Tie your account to your state identity. Let the platform know you. Let the government know you. Let the advertiser know you. Let the employer find you. Let every sentence become traceable, every confession recoverable, every political deviation attachable to a permanent record. This would be presented as safety. It would be sold as trust. It would be called accountability. But empires have always loved legibility. Bureaucracies love names. Police love maps of association. Corporations love identity graphs. Advertisers love verified targets. Employers love searchable obedience. Platforms love anything that turns the human being into a more stable unit of extraction. Every empire dreams of a world where every mask is removed except its own. A real-name internet would not abolish manipulation. Powerful actors could still buy speech, hire people, rent influence, create front organizations, operate through institutions, and launder propaganda through respectable channels. It would not end deceit. It would mostly make ordinary people easier to punish. Real-name internet solves the bot problem by wounding the human problem. It defeats artificial people by making real people more afraid. That is not a moral victory. The better distinction is not between anonymous and identified. It is between speech and reach. A person speaking under a pseudonym is one kind of act. A system manufacturing ten thousand pseudonyms to impersonate public opinion is another. A worker anonymously saying, “My company is lying,” is one thing. A corporation secretly funding a campaign of fake citizens to defend itself is another. A person writing a harsh review is one thing. A review farm flooding a marketplace with synthetic praise is another. A citizen criticizing a government is one thing. A state-sponsored swarm making that criticism disappear under waves of abuse is another. Speech is not the same as amplification. To speak is to offer a voice. To amplify artificially is to counterfeit a crowd. This distinction matters because freedom of speech has never meant the right to simulate the entire village. It has never meant the right to secretly buy the town square, hire actors to fill it, and then tell every passerby that “the people” have spoken. Free speech protects the person. It does not require society to accept forged evidence of mass agreement. Anonymous speech should be protected. Artificial reach should be accountable. That is the line. This is where a better internet might begin: not with a universal identity system, but with a contextual trust system. Ordinary speech should remain possible without papers. A person should be able to write, confess, criticize, explore, pray, grieve, rage, joke, and dissent without proving legal identity to the machine. But when speech is converted into power — when it becomes ranking, advertising, political influence, public metrics, reviews, petitions, fundraising, recommendation, mass commenting, or claims of consensus — stronger proof may be justified. The question should not be, “Who are you?” The question should be, “What kind of influence are you trying to exert?” If you want to post a poem under a false name, the internet should leave you alone. If you want to operate a thousand accounts to make your enemy appear hated by everyone, the internet should resist you. If you want to criticize your employer anonymously, the mask may be necessary. If your employer wants to create fake workers praising its own culture, the machinery should be exposed. If you want to say a politician is corrupt, you should not need to show your passport. If a campaign wants to purchase synthetic outrage and call it the voice of the people, it should be dragged into the light. One possible tool in such a system is anonymous personhood verification. The phrase sounds technical, but the idea is simple. It does not mean everyone must reveal their name online. It does not mean every website gets your passport. It does not mean the government should know every account you use. It means something narrower and more humane: You may not know who I am, but you can know I am one real human being. Imagine entering a theater. The usher does not need to know your mother’s maiden name, your politics, your employer, your medical history, your immigration status, or every theater you have ever attended. The usher only needs to know that your ticket is valid and has not already been used. Or imagine proving you are old enough to enter a place without handing every stranger a complete copy of your birth certificate. The claim matters; the whole identity does not. In a healthier internet, a person might be able to prove limited facts: I am a real human. I am not using this same human proof to create another verified account on this platform. I am old enough for this particular service. I am eligible to participate in this specific poll. I am a unique signer of this petition. I am not a swarm. But the platform would not necessarily learn the person’s legal name. The public would not see “John Smith from Austin.” The public might see only: verified human. Proof of personhood, not proof of name. The technical ethic is simple: Prove the minimum necessary fact and reveal nothing else. If the question is whether one real person is behind a petition signature, the system should not need to know that person’s employer, address, immigration status, family history, or full legal name. If the question is whether a review came from one unique human, the system should not need a permanent identity dossier. If the question is whether an account is part of a mass synthetic network, the answer should not require stripping every ordinary person naked before the platform. None of this is a magic solution. A verified human can still lie. A verified human can still be paid. A verified human can rent an account, sell a credential, join a brigade, repeat propaganda, or become the organic hand inside a synthetic campaign. Proof of personhood does not prove sincerity. It does not prove wisdom. It does not prove independence. It does not prove virtue. It only limits one form of fraud: the ability of one actor to cheaply become a crowd. That limitation matters, but it must not be confused with moral certification. “Verified human” does not mean trustworthy. It means only that a person, rather than an endlessly replicable machine identity, stands somewhere behind the act. The system must preserve that narrow meaning, or it will become another lie. A humane internet would not treat every act online the same way. It would have different trust requirements for different kinds of power. Ordinary anonymous speech should require no verification. People should be able to post, confess, criticize, explore, and dissent without proving legal identity. The frightened teenager, the closeted dissident, the undocumented worker, the abuse survivor, the person criticizing a boss, the person asking a shameful question, the artist trying on a voice, the addict telling the truth before he can bear to sign his name to it — these people should not have to pass through an identity gate to speak. High-reach distribution may require stronger trust signals. If an account wants major algorithmic reach, mass commenting power, trend-shaping influence, or repeated access to recommendation systems, platforms may reasonably ask for signs that the account is not part of a synthetic swarm. The issue is not whether the person may speak. The issue is whether the system should help that speech appear as mass public reality. Reviews, polls, petitions, fundraising, and marketplace ratings may require one-human-one-action protections. These systems claim to measure real human judgment. A product review is supposed to represent a customer or user, not a script. A petition is supposed to count supporters, not sockpuppets. A poll is supposed to capture people, not an army of throwaway accounts. When fake multiplicity corrupts the very purpose of the system, personhood verification becomes more defensible. Political ads and paid influence should require funding and provenance transparency. The public has a right to know who is buying persuasion. If money is being spent to shape political perception, the buyer should not be allowed to disappear behind the costume of spontaneous citizenship. Institutional speakers should face stronger disclosure rules than ordinary individuals. Corporations, governments, campaigns, lobbying groups, state-linked media, large advertisers, coordinated advocacy networks, influencer marketing operations, and AI content farms should not be able to move through the public square disguised as ordinary citizens. If an organization speaks, the public should know it is an organization. If a government speaks, the public should know it is a government. AI-generated mass content should be labeled and rate-limited when deployed at scale. The issue is not that a person used a tool to write a sentence. Human beings have always used tools. The issue is industrial synthetic speech: mass-produced content designed to impersonate human presence, flood discourse, manipulate ranking, or create the illusion of consensus. Platforms should be required to report what kind of traffic they are amplifying. Human, automated, paid, coordinated, institutional, synthetic, and unknown activity should not all be collapsed into one glowing number called engagement. A platform should not be allowed to sell a crowd without telling us how much of that crowd is real. This is not censorship. It is architecture. The system should not say, “You cannot speak unless we know who you are.” It should say, “You cannot secretly manufacture the appearance of a crowd.” Those are different moral universes. There is another reason platforms will resist this distinction: it threatens their economics. Platforms publicly hate bots, spam, scams, fake engagement, and coordinated manipulation. They issue reports. They announce enforcement actions. They remove networks. They condemn inauthentic behavior. They speak the language of integrity. But the deeper truth is more compromised. Many platforms profit from fog. Fake accounts can make a platform look alive. Fake engagement can increase time spent. Fake comments can create drama. Fake followers can flatter creators. Fake views can inflate inventory. Fake clicks can produce revenue. Fake outrage can keep people scrolling. Fake consensus can make content appear important. Fake activity can be sold, directly or indirectly, as attention. A platform built on engagement has a strange relationship with fraud. It is harmed by fraud when advertisers lose trust, users flee, regulators intervene, or scams become too visible. But it may benefit from fraud when the numbers go up: when activity looks abundant, the machine feels busy, investors see growth, advertisers buy impressions, creators chase metrics, and political actors pour money into influence. The platform does not always want to know too precisely how much of its life is real. A serious human-trust layer would force a brutal accounting. It would separate real human engagement from automated activity, paid activity, coordinated campaigns, institutional messaging, synthetic content, and unknown traffic. It would ask platforms to tell advertisers, users, regulators, and the public: this is human; this is machine; this is paid; this is organized; this is state-linked; this is unknown. Such clarity would make some numbers cleaner and smaller. That is why trust is economically dangerous. It does not merely remove fraud. It removes useful illusion. The question is whether a platform is selling human attention or the hallucination of human attention. If it is selling human attention, then verified humanity is valuable. Advertisers should pay more for real people than for ghosts. Marketplaces should value reviews from unique humans. Political systems should care whether apparent public opinion comes from citizens or scripts. Comment systems should rank real human participation above artificial flooding. Trust should become a premium. But if the business model depends on inflated scale, then accountable amplification is a threat. It says: count more honestly. Sell less fog. Stop calling every twitch of the machine a person. The economic stakes are therefore not secondary. They are central. A platform that distinguishes real human participation from synthetic activity is not only changing moderation. It is changing the price of attention. It is changing the value of influence. It is changing what “engagement” means. And that is why the solution cannot be left to platforms alone. The same companies that built vast systems to harvest attention cannot be trusted, by moral instinct alone, to measure the purity of that attention against their own interests. They need pressure, standards, law, competition, public scrutiny, independent audits, and cultural demand. Otherwise the phrase “verified human” will become another marketing badge, another trust costume, another way of selling the public a cleaner story about the same old machinery. The danger runs in the other direction too. A personhood system, if designed badly, could become monstrous. A hidden map could emerge: legal person to credential, credential to accounts, accounts to speech, speech to associations, associations to punishment. Even if the public sees only “verified human,” someone somewhere may hold the chain. A government may demand access. A corporation may monetize around it. A court may subpoena it. A hacker may steal it. An authoritarian regime may weaponize it. A future administration may reinterpret it. A platform may quietly use it for ranking, advertising, exclusion, and discipline. The surface may say anonymity. The basement may contain the registry. That is worse than honest identification because people may speak freely while falsely believing themselves protected. It is one thing to know you are naked. It is another to be told you are clothed while the cameras are already recording. A system built to prove humanity could become a system for licensing humanity. This is the knife edge. Verification can fight artificial crowds. It can also create a new gatekeeper over speech. It can protect trust. It can also produce a two-tier internet: verified people with reach, unverified people treated as suspicious noise. It can reduce bots. It can also exclude refugees, undocumented people, minors, the unhoused, people without stable documents, people in abusive households, people from sanctioned or unstable countries, people whose lives do not fit clean administrative categories. A trust layer can become a leash. And if it uses biometrics — eyes, faces, fingerprints, voices — the stakes become darker. Passwords can be changed. Documents can be reissued. But the body is not easily replaced. A leaked biometric system is not like a leaked password database. You cannot rotate your iris. You cannot patch your face. Even if a system claims to store no raw biometric data, the public must trust the hardware, the audits, the software, the incentives, the law, the issuer, the supply chain, and the future. That is a lot of trust to demand from people who already have good reasons to distrust institutions. Then there is function creep. A tool begins as optional protection against bot swarms. Then it becomes required for political comments. Then for videos. Then for payments. Then for job platforms. Then for news. Then for adult content. Then for encrypted messaging. Then for public services. Then, quietly, ordinary unverified speech still exists but is buried, downranked, demonetized, excluded from recommendations, treated as low-integrity by default. The right to post remains. The right to be seen disappears. This is how control often arrives in liberal systems: not as a ban, but as a ranking adjustment. So the safeguards cannot be decorative. They must be architectural. No single global identity provider. No universal mandatory credential. No platform access to legal identity for ordinary speech. No cross-platform tracking by default. No biometric monopoly. No use of personhood verification for behavioral advertising. No quiet downranking without transparency. No exclusion of people who lack conventional documents. No irreversible banishment without appeal. No deanonymization without serious due process. No system in which one corporation, one state, one protocol, or one vendor becomes the priesthood of human legitimacy. The cure for artificial people must not be a census of the soul. The future internet does not need to know everyone’s name. It needs to know when a crowd is real. It must protect the person who hides to tell the truth, and expose the machinery that hides to manufacture consensus. Protect the mask.Expose the machinery. And never mistake artificial noise for the voice of the people. —Elias WinterAuthor of Language Matters, a space for reflection on language, power, and decline This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit eliaswinter.substack.com [https://eliaswinter.substack.com?utm_medium=podcast&utm_campaign=CTA_1]
125 episodios
Comentarios
0Sé la primera persona en comentar
¡Regístrate ahora y únete a la comunidad de Language Matters Podcast!