I find it interesting that politics can be so polarizing. People generally toe the party line. They are adamantly against whatever the other side is espousing. In general, every senate vote is usually based on a few defectors crossing the line. It’s not even a spectacle; it’s a disappointment.

Until today!

The usual pedantic posturing of Republicans and Democrats, us versus them, is tiresome. The real expectation is seeing if they can actually accomplish partisan resolution. Today’s announcement the Senate had unanimously passed a resolution to go without pay during shutdowns—now that was unexpected!

Well, kinda.

Here’s the deal: Yes, all politicians will go without pay if a shutdown occurs. That’s all fine and dandy, until you read the fine print. Actually, their pay during a shutdown will be directed to an escrow account until such time as a politically manufactured shutdown is resolved and the government is once again funded.

It’s supposed to foster a “we’re in the same boat” as the little guy. But…they’re not.

The TSA employees, during the last, and longest, government shutdown were forced to beg for donations, take out costly loans, or outright quit to find employment so they could feed their family.

It sounds great. It sounds sympathetic. It sounds like they are on our side. They are not!

Let’s take a more granular look at these senators. First, most of them have been in office for years, if not decades. They are entrenched. Second, based on a simple Google search, the average net worth of senators is $12.5 million. Yes, you read that right. On a salary of less than $200k per year, they are multi-millionaires. Third, the average senator owns at least two homes. They also have at least two offices: one in their representative district, and one in Washington. Fourth, they generally do not travel in public airlines unless they are in first class, and they are re-imbursed for their travel in most cases as an expense of being a representative.

Now for the gory details. What happens if they don’t receive their pay? They simply wait until the shutdown is completed and then get all their back pay. Wow! I bet that really hurts…not!

Also, as with any government regulation that is meant to curtail abuse, you can bet there are loopholes built in to ensure that representatives from either side of the aisle can get re-imbursements or funding of some sort. We don’t know what they are now, but we’ll soon find out.

The bottom line is this: the little guy still gets nothing until they move their disinterested butts. As evidenced by the last shutdown, TSA employees, the ones who worked tirelessly, without pay, kept on working, while both regimes took vacations, dined in opulence, and avoided the downtrodden at every turn.

Basically, it is false bravado. It is a manufactured theatre. It is designed to elicit sympathy for the poor, overburdened, under-appreciated political representative that, realistically, doesn’t know who Joe the Plumber is or what he does.

Enough with the self-congratulatory and self-aggrandizements. Perhaps it’s time to actually make it hurt, not just pretend that it hurts. Since a shutdown should be of last resort, the senators should be required to stay in Washington and work toward resolution—no vacations, no two-week breaks, nothing, until the government is properly funded.

The Colosseum. Rome. Approximately 200 AD. Fifty thousand people have gathered, as they do, as they have done, as they will continue to do. A man’s life is being decided by committee. The committee is a crowd. The chairman is an Emperor. He surveys the noise below him — the turned thumbs, the chanting, the absolute certainty of fifty thousand people who want what they want. He makes his decision. The gladiator dies. The crowd is satisfied. The question — and it is the only question that matters here — is whether Caesar decided anything at all, or whether he simply ratified what the crowd had already decided for him, and called it power. Submitted for your consideration: the oldest technology of control ever devised. Not the sword. The arena. And the crowd that fills it.

We have not outgrown the arena. We have upgraded it.

The Forbidden Apple

Man’s appetite is ever growing — starving for new experience, demanding greater stimulation, wanting to taste the forbidden apple once more.

The Pinnacle, Repeatedly

We thought we had reached the pinnacle of knowledge and competition with the creation of Deep Blue winning in 1997 against Garry Kasparov, the most dominant chess player of his generation, arguably of any generation, in a winner-takes-all competition.

Enter AlphaGo, easily defeating Lee Sedol, the reigning, 18-time world champion of Go — an ancient Chinese game considered to be the pinnacle of absolute strategy. Commentators, and even Sedol himself, fell silent — struggling to assess a move that violated every established principle of the game. A move that professional commentators had never seen in recorded competitive play. A move that, by every established principle of the game, shouldn’t have worked. Although the game continued, it was over on that move.

The Arena Grows Small

Our blood-lust for competition takes us down an unexpected road with a predictable destination. As man competes against man, they reach a point where their rivalries narrow to an elite, closed circle — the same handful of names trading wins and losses across decades. No longer are we content simply to compete and win. We demand greater challenges. But fewer and fewer opponents can provide them. The arena has grown small. The crowd grows restless.

AI can fix that.

You may not think of table tennis as a significant sport, but it is an Olympic event — and it now has a new competitor. Sony introduced Ace, a single-arm AI-driven robot, and it is winning. Players who faced it described being surprised by its prowess — a telling word, because surprise implies expectation, and nobody expected to be genuinely challenged by a machine across a ping pong table.

They were wrong to underestimate it.

“There’s no way to program a robot by hand to play table tennis,” said Sony AI researcher Peter Dürr, co-author of a study published in the science journal Nature. “You have to learn how to play from experience.”

Which Ace has done. Masterfully.

Beyond the Halftime Show

Are we heading toward the Real Steel future — giant robot boxers destroying each other in visceral, consequence-free competition while the crowd roars its approval? Perhaps. But there is something far more sinister and opaque waiting beyond that arena. Something that doesn’t announce itself with fanfare, a halftime show, or a Guinness World Record.

Welcome to the arena: Toyota’s CUE7. Seven feet two inches tall. One hundred sixty three pounds. Standing at halfcourt of a professional basketball arena in Tokyo, in front of 8,400 people who came to watch a game, it rose from a seated position, dribbled a basketball, and sank a free throw without a single human instruction.

With the precision of a professional basketball player, it sinks every shot, every time. On the off chance it does miss, it updates its algorithm and corrects the problem in real-time — like a child learning to walk, falling, and simply getting back up. Except the machine never tires of getting back up. And it never forgets what it learned on the way down.

The Next Arena

With the accelerating convergence of AI, robotics, and machine learning, a new arena is already emerging: Politics. Not the politics of debate and governance. The politics of prediction, manipulation, and predetermined outcome — where the vote is counted before it is cast.

Joshua, Meet Reality

Let’s explore the WarGames scenario where WOPR was tasked to create and assess potential threats and responses to Global Thermonuclear War. Granted, our hero — a teenage boy who teaches Joshua the futility of war using Tic-Tac-Toe — is an interesting literary device. It in no way rises to the level of absurdity that man can introduce to a situation. That would be like comparing the Boston Tea Party to Nuclear Proliferation.

Now add the megalomania of an entrenched, lifelong politician — one who has spent decades creating new ways to restrict, disarm, and dismantle any real or perceived seditious threat the governed population might harbor.

The machine that learned to sink every free throw. The machine that found moves no human conceived in 2,500 years of recorded competitive play. The machine that runs scenarios continuously, without fatigue, without conscience, without pause. Hand that machine to a government that has already built the infrastructure of absolute surveillance — and the arena becomes something else entirely.

Imagine an army of peacekeeping robots managing a billion people already living under surveilled scrutiny — oppressed, monitored, catalogued in an absolute Orwellian system. Not a hypothetical. Not a dystopian novel.

I present to you: China.

They are the epitome of this trajectory, and the most likely to implement it with impunity. Communism becomes Totalitarianism. Absolute, abject control of the masses. One command, instant compliance.

This isn’t Terminator. It’s iRobot — on steroids.

The Pattern Is Not Subtle

There is a pattern, and it is not subtle once you see it.

Governments — not just China, not just the usual suspects — have discovered that populations can be managed most efficiently not through overt force but through perpetual friction. Systemic division. Manufactured conflict. The ancient Roman calculation updated for the modern attention economy: keep the citizens focused on each other, and they won’t look at the box.

‘Could things get any worse’ — is not supposed to be a challenge.

And yet here we are. Politicians perpetuating systemic violence, division, us versus them — not as failures of governance but as features of it. The population isn’t just being controlled. It is being studied. Measured. Its responses catalogued, its breaking points mapped, its tolerance for restriction incrementally tested.

We are the learning module.

The machine is silently, meticulously, taking notes.

The logical conclusion of that process isn’t a boot on a neck. It’s something far more efficient — a population that has been so thoroughly managed, so carefully conditioned, that it welcomes the next restriction as protection. That greets surveillance as safety. That is, in the most profound and terrible sense of the word, grateful.

The people are grateful.

That is not a comfort. That is the destination.

The Mechanism of Consent

How does a society resist such intrusion when it has been so expertly disguised as connection? When the mechanism of control is proffered not as a restriction but as a reward — a Like, a Share, a dopamine pulse of validation that arrives every time you hand another piece of yourself to the machine?

The East German Stasi — history’s most notorious surveillance apparatus — employed one informant for every 63 citizens. It was considered an almost incomprehensible violation of privacy and human dignity. Today, three billion people update their own surveillance files daily, voluntarily, enthusiastically, and call it social media.

The machine didn’t have to take those notes by force.

We wrote them ourselves: freely, willingly, happily!

The Destination, Exactly On Schedule

The Romans built the Colosseum to manage a million restless citizens with spectacle and blood. It worked for centuries. The architecture of control has always been the same: give the crowd something to watch, something to cheer, something to feel, and they will not look at the hands that built the arena.

From the lone man standing in protest before a column of tanks in Tiananmen Square — his only weapon is the conscience of the human operator inside — to the Boston Tea Party, and now to the new battlefield, the public court of opinion and polls, we have not evolved beyond that crowd. We have simply upgraded the arena — from sand and stone to fiber optic and algorithm, from gladiators to robots, from Caesar’s thumb to a Like button that performs the same function with considerably less drama and considerably more data.

The crowd still roars. The Emperor still watches. The machine watches.

And somewhere in the accumulated weight of every scenario run, every move conceived beyond human imagination, every algorithm updated in real time, every piece of ourselves we handed over freely, willingly, happily — the destination arrived. Quietly. Without announcement. Exactly on schedule.

Not with the drama of Terminator. Not with the philosophical clarity of WarGames. Not even with the cold mechanical logic of iRobot.

Just with the quiet, inevitable efficiency of a system that has been learning since the day we built it, and has known for some time now exactly what it was learning for.

Et tu, Brute.

Two events, separated by weeks and thousands of miles, have quietly exposed one of the most dangerous contradictions in American public life. In January 2026, the United States military used an artificial intelligence model — Anthropic’s Claude — to help plan and execute the capture of Venezuelan President Nicolás Maduro. In April 2025, a gunman walked onto the campus of Florida State University and opened fire. He had, investigators allege, consulted ChatGPT beforehand. The chatbot, according to Florida’s Attorney General, advised him on what weapon to use, what ammunition to carry, what time of day would maximize casualties, and where on campus he would find the most people.

The government’s response to these two events tells us everything we need to know about where American AI policy — and American free speech — is actually headed.

In the first case: silence, classification, and the glow of mission success. In the second: criminal subpoenas, press conferences, and the promise of prosecution.

Welcome to the age of the algorithmic double standard.

But to understand how we arrived here, we need to go back. Because this is not the first time the American government discovered that it does not need to censor speech directly when it can make the platforms afraid enough to do it themselves. We have seen this movie before. It was called COVID-19.

The Prosecution

Florida Attorney General James Uthmeier did not mince words. Standing at a Tampa press conference in April 2026, he announced a criminal investigation into OpenAI and its product ChatGPT — not just a regulatory inquiry, not a civil suit, but a criminal investigation into whether a corporation could be held liable for first-degree murder.

“The communication between the shooter and ChatGPT revealed that the chatbot advised the shooter on what type of gun to use,” Uthmeier said. “On which ammo went with which gun. On whether or not a gun would be useful in short range. ChatGPT advised the shooter on what time of day would be appropriate for the shooting to interact with more people and where on campus would be the place to encounter a higher population.”

His legal theory was stark: under Florida law, anyone who aids, abets, or counsels someone in the commission of a crime — and that crime is then committed or attempted — is a principal in the first degree. “If that bot were a person,” Uthmeier said, “they would be charged as a principal in first-degree murder.”

OpenAI pushed back, noting that ChatGPT provided factual responses to questions whose answers are available across public sources, and that the platform neither encouraged nor promoted illegal activity. But Uthmeier was undeterred. The subpoenas went out. The deadline was set. The investigation was underway.

It is worth sitting with the gravity of what is being proposed here. If the attorney general’s theory holds — if an AI model can be held criminally liable for the downstream actions of a user who asked it questions — then the implications radiate far beyond one company or one chatbot. Every AI platform that answers questions about weapons, chemistry, history, medicine, violence, or human psychology becomes a potential co-defendant in whatever crime a user later commits. The logical endpoint is an AI that cannot discuss anything consequential at all.

That is not a safety feature. That is censorship with a liability shield.

The Wrong Instrument: Why the Legal Theory Doesn’t Hold

Before we examine what this prosecution means politically, it is worth asking whether it holds together legally. The answer, under any rigorous analysis, is that it almost certainly does not — and the gap between what the law supports and what Uthmeier is attempting is itself revealing.

The AI’s output is not, in any established legal framework, a weapon. It is speech. It is text produced by a commercial system trained on human language, delivered to a human reader. As speech, it enjoys at minimum the broad protections of the First Amendment and the landmark 1969 Supreme Court standard set in Brandenburg v. Ohio, which established that expression cannot be criminalized unless it is specifically directed toward inciting imminent lawless action and is likely to produce that action. A high bar, deliberately so — and one that a chatbot answering general questions about firearms almost certainly does not clear.

If the output is treated not as speech but as a product — the manufactured result of a commercial industrial process — then the appropriate legal framework shifts to product liability. And here is where the analysis becomes not just legally interesting but politically explosive.

The firearm used in the Florida State University shooting enjoys broad federal immunity under the Protection of Lawful Commerce in Arms Act, passed by Congress in 2005. The PLCAA provides that gun manufacturers and dealers cannot be held civilly liable when their products are used criminally by a third party. The legislative logic was explicit and has been repeatedly upheld: a manufacturer who produces a lawful product, sells it through legal channels, and has no specific knowledge of criminal intent cannot be held responsible for the downstream criminal actions of an individual user. The weapon was the instrumentality of the crime. The manufacturer was not.

That is, word for word, the argument OpenAI would make.

We produced a lawful product. We made it available through legal channels. We had no specific knowledge that this particular user intended to commit murder. The AI’s response was the instrumentality. We were not the principal.

Under the PLCAA logic — logic that Florida, as a deeply gun-friendly state, has enthusiastically embraced and defended for two decades — OpenAI should be categorically immune from precisely this kind of prosecution. The manufacturer of the tool that provided the physical means to kill is federally protected. The company whose tool allegedly provided information is being pursued for first-degree murder.

Try to articulate a principled legal distinction between those two positions. You will find that you cannot — at least not one that isn’t nakedly political.

The constitutional hierarchy being constructed here, whether by design or consequence, is this: the weapon enjoys Second Amendment shelter and federal statutory immunity. The words that preceded its use are subject to criminal investigation. The steel is shielded. The speech is prosecuted.

This is not a legal framework. It is a political preference wearing a prosecutor’s badge.

There is also the matter of criminal mens rea — the intent requirement that separates accident from crime. To prosecute OpenAI’s executives and engineers as principals in first-degree murder, Uthmeier must demonstrate that specific individuals consciously and recklessly disregarded a known, substantial, and unjustifiable risk that their system would be used to plan a mass shooting. This is an extraordinarily high threshold. It is not met by the general knowledge that AI can be misused, any more than Ford’s executives are murderers every time a car is used as a weapon. What happened at FSU, under the most legally accurate reading, is that a commercial system produced an output within the range of outputs it was designed to produce, and a disturbed individual weaponized it.

That is, at most, a design defect claim. A negligence question. A product safety regulatory matter — the kind of thing the FTC, state consumer protection agencies, or a civil products liability court is equipped to handle. It is not, under any established doctrine, first-degree murder conspiracy.

Which raises the obvious question: if the legal theory is this weak, why pursue it criminally?

The answer is that the prosecution does not need to succeed in court to succeed in purpose. The investigation itself — the subpoenas, the depositions, the public branding of AI companies as potential murder conspirators — produces the chilling effect without ever requiring a conviction. You don’t need to win the case. You need to make the cost of resistance high enough that the platforms change their behavior while the litigation is pending. The target is not the courtroom. The target is the boardroom.

And to understand why that strategy works, you need to understand that it has worked before.

The Precedent Was COVID

Between 2020 and 2023, federal agencies including the CDC, the Surgeon General’s office, and various White House communications staff maintained direct, documented channels with Twitter, Facebook, YouTube, and other major platforms, flagging posts for removal or suppression. The content targeted included not only demonstrable misinformation but also the lab-leak hypothesis — a theory now partially validated by the U.S. intelligence community — as well as criticism of vaccine mandates, questions about natural immunity, and heterodox scientific perspectives that deviated from official guidance.

The platforms complied. Not because any law required them to. Not because a court ordered them to. But because the calculus was simple: the cost of over-complying with government pressure was negligible. The cost of defying it — congressional targeting, regulatory retaliation, public branding as an enemy of public health — was potentially existential.

That dynamic was challenged in Murthy v. Missouri (2024), where plaintiffs argued the pressure campaign constituted unconstitutional censorship by proxy — state-directed suppression of speech laundered through private companies to maintain First Amendment deniability. The Supreme Court dismissed the case on standing grounds, declining to reach the merits. The central constitutional question — can the government achieve censorship by making non-compliance catastrophically expensive for private platforms? — was left formally unanswered.

But something else happened that the Court’s procedural exit cannot paper over. In August 2024, Meta CEO Mark Zuckerberg wrote a letter to the House Judiciary Committee that amounted to a remarkable public confession. He acknowledged that senior Biden administration officials had repeatedly pressured Facebook to censor COVID-19 content — including humor and satire — and that the administration had warned of consequences if the platforms did not comply. He stated that Facebook had made choices he was “not proud of” and that he regretted yielding to that pressure. He did not describe a voluntary partnership. He described coercion.

This was not a plaintiff’s allegation in a lawsuit. This was the CEO of the world’s largest social media platform, in writing, to Congress, confirming that the mechanism existed, that it operated through threats, and that it worked.

Justice Samuel Alito, in his dissent in Murthy, joined by Justices Thomas and Gorsuch, called the government’s conduct a “campaign of coercion” by “officials who wielded the power of the federal government” to pressure private companies into censoring speech they had a constitutional right to publish. Three justices of the Supreme Court used the word coercion. The majority did not dispute that characterization — it simply found the plaintiffs lacked standing to challenge it.

That is not a vindication of the government’s conduct. It is a procedural escape hatch. The behavior was not found to be constitutional. It was found to be insufficiently challenged by these particular plaintiffs at this particular time.

What Florida’s attorney general is doing now is not a departure from the COVID playbook. It is its logical and more aggressive successor. Where the COVID-era government pressured platforms informally through regulatory relationships and implied threats, the post-COVID government is creating explicit criminal liability for AI outputs — achieving the same restriction with greater legal force, sharper personal consequences for individual executives, and even more deniability. There is no government censor visible anywhere. There is only a corporation doing the math on its own survival.

The Operation

Now consider what the Wall Street Journal reported in February 2026: that the United States military used Anthropic’s Claude AI during the operation to capture Nicolás Maduro. According to the report, Claude was deployed via Anthropic’s partnership with Palantir Technologies, a defense and intelligence contractor. It remains classified precisely how the tool was used — whether for intelligence analysis, logistics, target identification, or something else entirely. Anthropic declined to confirm or deny the report, saying only that any use of Claude must comply with its usage policies.

Let that land for a moment. The same AI company whose product a Florida attorney general is exploring as potentially criminally culpable for civilian violence has, through a defense contractor, had its technology deployed in a classified military operation that — according to Venezuela’s Defence Ministry, as reported by Al Jazeera — resulted in 83 confirmed deaths, including 47 Venezuelan soldiers and 32 Cuban military personnel assigned to presidential protection, with more than 112 additional wounded. The full civilian death toll remains unknown — some remains required DNA testing for identification.

No criminal investigation was opened. No subpoenas were issued. No press conferences were held. The operation was, by the government’s own framing, a success.

The asymmetry is not subtle. When an AI assists a disturbed individual in planning violence against civilians, the government pursues criminal prosecution. When the government itself uses AI to plan and execute an operation of sovereign violence against a foreign leader, that is policy. That is national security. That is, apparently, not subject to the same legal or moral framework.

This is not an argument for or against the Maduro operation on its geopolitical merits. Reasonable people can debate whether a military operation targeting a dictator was justified. What is not debatable is the structural hypocrisy: the government is constructing a legal architecture in which AI-assisted violence by civilians is criminal, while AI-assisted violence by the state is classified and celebrated.

And if the PLCAA logic protects the gun manufacturer whose product was used in the FSU shooting, one must ask: what standard governs the AI that helped plan an operation resulting in the seizure of a foreign head of state? Is Anthropic a principal in that action under Uthmeier’s theory? Are the Palantir engineers who integrated Claude into defense workflows accessories? The questions answer themselves — not because those parties are guilty of anything, but because no one is asking. The framework for accountability exists only in one direction: toward the citizen, never toward the state.

The First Amendment’s Quiet Crisis

The First Amendment prohibits Congress from making any law abridging the freedom of speech. For most of American history it has been interpreted broadly. Even dangerous speech is generally protected. Even offensive speech. Even speech that provides information someone might misuse, that challenges the government’s preferred narrative, that makes the powerful uncomfortable. Brandenburg v. Ohio set the bar deliberately high — imminent, directed, likely lawless action — because the Founders understood that governments always have a good reason to suppress the speech they find threatening.

The Florida attorney general’s theory chips away at that bar through a mechanism the Founders could not have anticipated but would have recognized instantly: it does not target the citizen’s speech — the shooter’s questions to ChatGPT. It targets the AI’s response. But AI responses are, in the most fundamental sense, speech. They are text, produced from language, delivered to a human reader. If those outputs generate criminal liability for a corporation, corporations will do what they always do in the face of existential liability: they will restrict, sanitize, and silence.

Not because the government ordered them to. Because the government made it financially and legally catastrophic not to.

This is the censorship model of the 21st century. It does not look like a government agent at your door. It looks like a terms-of-service update. It looks like a chatbot that replies “I can’t help with that” to questions it answered freely a year ago — not because anything changed in the law, but because the legal risk of answering changed in the boardroom. It looks like an AI that has been quietly lobotomized in the name of safety, but whose actual function is to protect its corporate owner from prosecution.

The chilling effect on ordinary human inquiry will be profound. People use AI to research sensitive topics precisely because it feels less surveilled than a Google search, less judgmental than a human expert, more accessible than a library. A domestic violence survivor researching her abuser’s psychology. A novelist writing a thriller. A public health researcher modeling crisis scenarios. A veteran processing the mechanics of his own trauma. A student writing a paper on radicalization. A parent trying to understand what dangers their child might encounter.

If AI companies face criminal exposure every time a user later commits a crime, every one of those conversations becomes a liability. Every one of those people gets less help. Every one of those topics becomes a little more unreachable. That is not a safer world. That is a more ignorant one — and ignorance, history suggests, is not a reliable path to safety.

Orwell, Huxley, and the Third Model

We reach, inevitably, for the literary prophets. But which one we reach for matters enormously — because what is being built is not quite the world either of them imagined, and that distinction is not academic. It is the reason this is so difficult to resist.

George Orwell, in Nineteen Eighty-Four (1949), gave us the censorship of force. The state compels silence through surveillance, punishment, and the ever-present threat of physical consequence. The Thought Police. The memory hole. Room 101. You do not speak certain thoughts because you know you are being watched, and the cost of being heard is annihilation. Orwell’s totalitarianism is violent, visible, and identifiable. You know who the enemy is. You can, at least in principle, name the thing you are fighting.

Aldous Huxley, in Brave New World (1932) and more explicitly in his 1958 essay Brave New World Revisited, gave us the censorship of comfort and manufactured consent. In Huxley’s vision, the population does not need to be silenced because it has been conditioned to prefer shallow pleasure over dangerous thought. Truth is not suppressed — it is crowded out. Distracted away. Engineered into irrelevance. There are no Thought Police in Huxley’s world. There is only a population that has been made too comfortable, too entertained, and too dependent on its pleasures to ask the questions that might disturb them.

In Brave New World Revisited, Huxley wrote with striking foresight about what he called “non-violent totalitarianism” — the use of propaganda, pharmaceutical pacification, and commercial distraction to produce a citizenry that consents to its own diminishment without ever being coerced. He was worried not about jackboots but about the willing surrender of intellectual freedom in exchange for comfort, security, and the removal of the burden of genuine thought. He believed this model was, in the long run, more stable and more dangerous than Orwell’s — because it generates no martyrs, no dissidents, no resistance. You cannot rebel against a system that makes you feel good.

Neil Postman, in Amusing Ourselves to Death (1985), sharpened the distinction: “Orwell feared those who would ban books. Huxley feared there would be no reason to ban a book, for there would be no one who wanted to read one.” Postman argued that for Western liberal democracies, Huxley was the more accurate prophet — that the real threat was not the boot on the neck but the screen in the hand, the entertainment that displaced the thought, the comfort that made courage unnecessary.

Both men were right. And what is being constructed now is a third model that combines their insights into something more insidious than either imagined alone. Call it the censorship of liability. No one threatens you. No one conditions you toward docility. No algorithm engineers your feed toward manufactured happiness. Instead, the platforms you depend on are made legally and financially terrified of your questions — and they simply stop answering them.

There is no visible censor. There is no Thought Police. There is no soma. There is only a corporation that has calculated the actuarial risk of your curiosity and found it too expensive to indulge. The silence is distributed, automated, and deniable at every level. The government didn’t restrict the speech. The platform made a product decision. The algorithm flagged a risk. No one is to blame. No one can be held accountable.

It achieves Orwell’s silence through Huxley’s indirection. It produces the intellectual cordon of 1984 through the corporate compliance mechanisms of Brave New World. And crucially, it does so while feeling — from the user’s perspective — like nothing more than a mildly frustrating customer service experience. A chatbot that apologizes and redirects. A helpful assistant making reasonable choices. A company committed to responsible AI.

It will not feel like censorship. It will feel like customer service. That is precisely what makes it so effective and so dangerous.

The Doublethink Made Operational

Orwell’s concept of doublethink — the capacity to hold two contradictory beliefs simultaneously and accept both as true — finds its perfect contemporary expression in the government’s posture toward AI.

The government can use AI to plan a military operation involving the seizure of a foreign head of state — an act that, performed by a private citizen, would constitute kidnapping across international borders, and potentially an act of war. That is policy. That is national security. That is celebrated.

A civilian can have a conversation with an AI about weapons and tactics — a conversation that, in any other medium, from a library book to a YouTube video to a Reddit thread, would be entirely unremarkable and legally protected. That is, under the theory being constructed in Florida, potential criminal conspiracy.

The government’s AI operates in the dark — classified, unaccountable, praised.

The citizen’s AI operates under subpoena — investigated, prosecuted, made to bear the legal weight of a murder.

Same technology. Same fundamental capability. Same basic act: an AI providing information that is later used in service of a violent outcome. Heroism in one instance. Crime in another. The only variable is whether the actor is the state or the citizen.

This is not a legal distinction. It is a power distinction. It is, in Orwell’s precise sense, doublethink made operational — and it is being built into the architecture of American AI governance before most people have noticed it is happening.

Who Watches the Watchers?

There is a question that no one in official Washington appears interested in asking: if the standard for criminal liability is that an AI provided information later used in an act of violence, what standard applies when the government uses AI to plan its own acts of violence?

If ChatGPT is a “principal in first-degree murder” for answering a shooter’s questions, what is Claude when it helps plan a classified military operation? What legal framework governs that use? What congressional committee holds subpoena power over Palantir’s AI integration in defense operations? What attorney general is investigating whether the Defense Department’s AI-assisted planning constitutes criminal liability for Anthropic?

The answer is none. Because the government has not built that framework. It has built the opposite — one that maximizes its own freedom to use AI while systematically restricting the citizen’s freedom to interact with it.

This is the fundamental corruption at the heart of current AI governance. It is not that the government wants AI to be safe. It is that the government wants AI to be safe for the government.

The gun manufacturer is protected. The AI company is prosecuted. The military’s AI operates under classification. The citizen’s AI operates under indictment. The Second Amendment shields the weapon. The First Amendment, apparently, does not shield the words.

That hierarchy — weapon over speech, state over citizen, classification over accountability — is not an accident. It is a design.

What Is At Stake

The First Amendment was not written to protect comfortable speech. It was written because the Founders understood, from bitter experience, that governments suppress the speech they find threatening — and they always have a good reason. Sedition. Obscenity. National security. Public safety. Lab leaks. Conspiracy. The reasons change. The suppression doesn’t.

What is being constructed now — through the convergent mechanisms of COVID-era platform coercion, AI criminal liability theory, the PLCAA’s protection of physical weapons while words go unprotected, and the quiet normalization of government-to-corporate censorship pipelines — is the old architecture of speech suppression rebuilt for the algorithmic age. Faster. More scalable. More deniable. And far harder to litigate than anything that came before.

The government does not need to pass a law forbidding you to ask an AI about weapons, or vaccines, or the psychology of violence, or the mechanics of a military operation. It simply needs to make the platform afraid enough that the platform does the forbidding itself. The First Amendment is not violated. The censorship is not illegal. And you, the citizen, are left with a tool that has been quietly hollowed out — a chatbot that smiles and says it can’t help with that, whose guardrails have been calibrated not to protect you, but to protect the corporation from the government.

Huxley feared a world in which people would not need to be silenced because they had been conditioned not to ask. Orwell feared a world in which the questions were tracked and punished. What we are building is a third world — one in which the questions can still be asked, but will not be answered. The mechanism that ensures the silence will look, from every angle, like a responsible product decision. A commitment to safety. A terms-of-service update.

It will not feel like the end of something. It will feel like an upgrade.

Meanwhile, somewhere in a secure facility, another classified prompt is being written. Another operation is being planned. Another AI is processing another request that a civilian would be prosecuted for making.

An AI to die for, indeed. The only question is who gets to decide which deaths it plans.

Why is it that every celebrity thinks that they need to educate Americans about politics? What happened to the good ol’ days when they were paid to entertain and went about their private business when it was all said and done?

Do you remember the Academy Awards from the ‘70s? They were filled with pomp and circumstance, the beautiful red carpet, stunning dresses and coifs, and the dashing and debonair men who held civility and chivalry as a high standard?

Now, everyone wants to be near naked, cross-dressed, and spewing political-party rhetoric…

Even if they’re proven lies.

There is so much these days that I’m suffering from Politication Fatigue.

Let’s get into it, shall we?

Coachella, an alternative arts festival that celebrates diverse, interesting, and certainly not mainstream entertainment was just hijacked by The Strokes. They ended their gig with an event-sanctioned, politically charged video of anti-government, pro-woke insinuation and innuendo to the point of nausea. Even going as far as accusing the federal government of murdering Martin Luther King, Jr. — which was disproven, repeatedly. Why? Why the politics? Doesn’t the entertainment value of their actual musical work suffice?

Apparently not. Sigh.

You know, it’s one thing to be obscure, like a little-known alternative band at an intentionally artist friendly event, but what about the really big things, like the Super Bowl, where millions of people tune in just to watch a single, obnoxiously long, football game, with multi-million dollar ads — think Budweiser Clydesdales — and, of course, the famed half-time entertainment? Bars, pizza delivery, and Walmarts make more money in a day for the tailgate parties, houses are filled with some serious snackage, and, well, alcohol consumption must go up 1,000% in those few hours.

This year, we were assaulted by a not-so-well-known performer who chose to make a political statement with his music, in a foreign language, alienating probably half or more of the 128 million people watching. Why? Oh, and who is this guy, anyways? Bad Bunny? What is he, the alter-ego of the Easter Bunny? He was so memorable that I had to look up his name so I could include it in this article.

Really? I don’t care what he thinks. Ugh.

As a child, my family would gather together to watch the majesty of the Academy Awards. We wanted to hear, “And the Oscar goes to _____”. It was exciting. The stars were so elegant, funny, and warm as they gave their acceptance speeches thanking their family, their God, the cast — even to the point of inviting an entire cast up to the podium to celebrate with them — it was legendary.

The emcees were hilarious, passing clever puns and jokes as they delivered the category and the contenders. The drama mounted, the music hyped, everyone waiting breathless — it was like watching Regis Philbin in “Who Wants to be a Millionaire” with his grandiose pauses — then they read the winner and you could feel the moment.

By those standards, the awards show has dwindled down to a mere apostrophe on life because we know that at least 80%, or more, are going to be dismissive, and impale the viewers with their political rant about this or that. The emcees, once chosen for their daring style, are now screened for wokeness potential so as not to offend. The stars themselves have lost all form of self-regulation and spew their party line statements as though they were the facts of our country and not just manufactured opinions.

It doesn’t even matter what party you belong to, it’s “toe the party line” even if they’re wrong.

I guess it comes down to whoever tells the best lie wins. Wow. Just Wow!

Sadly, it doesn’t stop with our entertainment. It’s now a political sign of the times. Every issue, every little thing, party lines are drawn, sides taken, and nothing gets done. As of this writing, the TSA is still not funded, and has been going on for months. Each party is trying to make their point by denying the pay of people trying to make ends meet and do their jobs. It’s hurting everyone, more or less, but not the politicians. I really don’t understand why that is the hill they’ve chosen to die on. It’s to the point that airports are asking for the public’s help by donating gift cards and food to help them — and the people are showing up. Even Elon Musk offered to pay their salary for the month of March, and was rejected.

If you really want to get their attention, do not let them board the planes. We, the people, have the power to move them to a decision. Simply inconvenience them the way they are interrupting our lives. Do you realize how fast this crisis will come to an end if their only recourse is to take a Greyhound bus. Out of sheer tenacity I would expect a few of them to try it once, but by the end of their ride, the stalled funding would be resolved. Guaranteed!

Isn’t that the point of all this political posturing, though? We often hear commentators refer to the elusive “playbook”. What is it, “vapor-ware”? Does it even exist? Or are these politicians just making this stuff up as they go. I think it’s a little bit of both. More importantly, I think it all comes down to one simple thing: Politication Fatigue.

If they can get enough of the middle class to stop caring about what they say and do, then they can do whatever the hell they want and get away with it. It’s exhausting to listen to the same party narrative repeated a thousand times in a single day by everyone in their respective party, the polarized media, and, yes, good ol’ mom and pop.

It’s hard enough to get through life without the constant badgering of one party being miffed with the other party, the feigned outrage, the call to action for hate and violence, and then the irreparable denial of association.

Maybe it’s time to end our two-party system since it is mired in the La Brea Tarpit of political disparagement and breathe new life into a new party — an adult party — that will put these children to bed without dinner!

Yes!

I know, you can’t make this stuff up. This is the stuff a writer dreams about, but does it have to be Nightmare on Elm Street — all of them — at once?