In 2024, the University of Pennsylvania’s Center for Ethics and the Rule of Law ran a series of high‑level U.S. civil conflict simulations under Professor Claire Finkelstein, mapping how a standoff between federal authorities and a defiant governor might spiral into something closer to civil war. In January 2026, the initial stages of that scenario appeared on American streets, as the confrontation now tearing through Minnesota began to mirror the exercise’s opening moves. An academic “what if” had become a live‑fire dress rehearsal.

They have succeeded. But in doing so, they have also legitimized a dangerous idea: that reality itself can be modeled, gamed, and won. What started as an educational experiment in controlled environments has escaped the campus. In the private sector, it has found a very different purpose.

From the Classroom to the Gray Zone

Enter Crix22, a theatrical simulation company built on world models. It is not an educational institution. According to its website, it builds simulations for corporations, universities, and even museums. Buried in the fine print at the bottom of the site is a product called Intelligence Lab, restricted to those with a SAM.gov account. That is not typical for a business simulation company.

Where Wharton’s simulations are for a grade, Crix22’s are for a client. And their results are not hypothetical.

On December 19, 2025, the investigative outlet Drop Site News published a bombshell report linking Jeffrey Epstein to the murky afterlife of the Iran‑Contra scandal. In the days that followed, according to an anonymous source with the buyer of the simulation, Crix22 was contracted to run a rapid‑response simulation on the likely outcomes of this revelation. The scenario was codenamed DROP‑EPSTEIN. Among its most striking predictions: that the House of Windsor would collapse as a constitutional institution by the year 2030.

Now they haven’t been proven correct yet, but how does one get from Epstein and the Iran Contra Affair to the fall of the British Monarchy.

On February 19, 2026, Prince Andrew was arrested.

This is the crucial difference. Wharton teaches the theory; Crix22 sells the application. The firm’s methodology isn’t magic. It is fast, ruthless diagnosis. It appears to have taken publicly available intelligence, processed its structural implications faster than almost anyone else, and modeled the most probable cascade. Then, allegedly, it sold that model.

The Arms Dealer’s Defense

The danger of this new industry is not that the models might be wrong, but that they might be right. A simulation that can accurately map the failure points of a government or a global corporation is not an analytical tool; it is an operational playbook. It is a weapon.

Crix22 ignored repeated requests for comment. Eventually, an attorney for the firm replied by email, stating that Crix22 has no knowledge of any scenario codenamed DROP‑EPSTEIN, that its simulations are built for “training and academic purposes only,” and that any other insinuation would be grounds for a defamation claim. That was quite a response.

This is the arms dealer’s creed, updated for the 21st century: the maker of the weapon declares itself neutral, even as it profits from instability. While UPenn can shelter behind the noble cause of education, Crix22 operates in the gray zone of the global intelligence market, where models are not used to teach future leaders how the world works, but to give an edge to the highest bidder.

As one prominent defense‑tech founder recently argued in a New York Times essay about “our Oppenheimer moment,” the race to build AI weapons is already underway; hesitation will be punished. Crix22 operates downstream from that logic, but in a different market: not building the weapons, but mapping the worlds they might be used in.

CARACAS-SHIFT

The leaked cache that described DROP‑EPSTEIN also contained references to CARACAS‑SHIFT, a July 2025 simulation that predicted the removal of Nicolás Maduro. CRIX22’s counsel acknowledged that the company’s model assigned a high probability to a post‑Maduro transition in Caracas and a U.S. move on Venezuela’s Orinoco Belt oil fields. “As did many analysts at the time,” he noted, emphasizing that CARACAS‑SHIFT was intended as a “purely hypothetical exercise.”

The Pattern, Not the Coincidence

This is the new, unregulated reality. The ivory tower has, with the best of intentions, laid the intellectual groundwork for a new form of power. Now, private, unaccountable firms are selling that power to whoever can afford it. We are not at the beginning of a new cold war. We are already in the middle of a new, very hot, and very private one, fought with models and simulations instead of soldiers and spies. And the people who wrote the rules of the game are still in the classroom, teaching.

As a new generation of Iranians chant the name of his long-dead father in the streets, the story of the Pahlavi dynasty — its dramatic rise, its brutal reign, and its spectacular fall — has become essential reading. It is a story of two Shahs: a father who built a nation from dust, and a son who lost it all to revolution.

The Father: A Soldier Forges a Nation

The Pahlavi story begins not in a palace, but in a village. Reza Khan, born in 1878, was a common soldier of obscure origin who rose through the ranks of the Persian Cossack Brigade, a Russian-led military unit. In 1921, with the backing of the British who saw him as a bulwark against the Bolsheviks, he marched his troops on Tehran and seized power from the crumbling Qajar dynasty.

By 1925, he had deposed the last Qajar shah and crowned himself Reza Shah Pahlavi, taking a name that deliberately evoked the pre-Islamic glory of the Sasanian Empire. He was an autocrat, a modernizer, and a nationalist in the mold of his contemporary, Turkey’s Mustafa Kemal Atatürk. He built the Trans-Iranian Railway, established the University of Tehran, banned the veil, and forced men into Western suits. He was ruthless, crushing tribal rebellions and centralizing power with an iron fist. But his flirtation with Nazi Germany proved his undoing. In 1941, fearing he would side with the Axis, the Allies (Britain and the Soviet Union) invaded Iran and forced him to abdicate in favor of his 22-year-old son, Mohammad Reza. Reza Shah died in exile in South Africa in 1944, never to see Iran again.

The Son: The King of Kings and the Revolution

Mohammad Reza Pahlavi began his reign as a constitutional monarch, but the 1953 CIA-backed coup that overthrew his popular Prime Minister, Mohammad Mossadegh, changed everything. After being briefly forced to flee, the Shah returned to power as an absolute ruler. For the next 25 years, he ruled Iran with a combination of grand ambition and brutal repression.

His “White Revolution” in 1963 brought land reform, women’s suffrage, and a massive literacy campaign. He built a modern military, a network of highways, and a booming economy fueled by oil wealth. But beneath the veneer of progress was the iron hand of SAVAK, his notorious secret police, trained by the CIA and Israel’s Mossad. Thousands of dissidents were tortured and executed. The Shah’s lavish 1971 celebration of the 2,500th anniversary of the Persian Empire at Persepolis, costing hundreds of millions of dollars, became a symbol of his detachment from the lives of ordinary Iranians.

By the late 1970s, a perfect storm was brewing. Economic recession, combined with the Shah’s increasingly autocratic rule and the growing influence of an exiled cleric named Ayatollah Ruhollah Khomeini, created a revolutionary movement. The Shah, secretly battling cancer and seemingly paralyzed by indecision, failed to grasp the scale of the opposition until it was too late. On January 16, 1979, he left Iran for what was officially described as a “vacation.” He would never return.

The Exile: A 47-Year Wait

The Shah died of lymphoma in Egypt in 1980, a broken man. His son, Reza Pahlavi, who was training as a fighter pilot in the United States when the revolution occurred, became the pretender to the Peacock Throne. For nearly five decades, he has lived a life in exile, a king without a kingdom.

Now, as the Islamic Republic faces its own existential crisis, the ghost of the Shah has returned. The chants of “Javid Shah!” on the streets of Tehran are not necessarily a call for the restoration of the monarchy, but a rejection of the 45-year rule of the clerics. They are a cry for the Iran that was lost — a modern, secular, and prosperous nation.

Reza Pahlavi, a man who has spent his entire adult life waiting for this moment, now finds himself at the center of a complex and dangerous game. He is a symbol of a past that many Iranians idealize, but also a reminder of an autocracy they do not wish to repeat. Whether he can unite the fractured opposition and offer a credible path forward remains the great unanswered question of Iran’s future. The soldier’s son lost a throne; can the Shah’s son reclaim it?

Seventy-five years later, as we look back from 2026, that broadcast remains the most significant moment in the history of “spectacle warfare.” It was the day the apocalypse became a household guest.

The Engineering Madness of Klaus Landsberg

Broadcasting live from the Nevada desert in 1951 wasn’t as simple as hitting “Go Live” on a smartphone. It was a logistical nightmare. Klaus Landsberg, the visionary behind Los Angeles station KTLA, had to bridge a 300-mile gap between the Nevada Test Site and the Hollywood hills.

• The Relay: He built a “daisy chain” of microwave transmitters.
• The Peaks: Engineers hauled massive, sensitive equipment up to the freezing, wind-swept summits of Mt. Charleston and Mt. San Antonio.
• The Signal: They literally bounced the image of a nuclear explosion from mountain peak to mountain peak until it hit the KTLA transmitter.

When the countdown hit zero, the signal held. For the first time, a civilian audience saw the “flash” in real-time.

News Nob: The World’s Deadliest Press Box

The reporters weren’t tucked away in a safe studio. They were stationed on a ridge 10 miles from ground zero, famously nicknamed “News Nob.” Reporters like Bill Welsh stood in the biting desert wind, wearing little more than trench coats and fedoras, staring into the abyss. When the bomb—a 1-kiloton “Operation Ranger” shot—detonated, the thermal pulse was so intense that the KTLA cameras were momentarily “blinded.” The screen went white, a haunting visual metaphor for the destructive power of the weapon.

“It was as if the sun had decided to rise at 5:45 in the morning, and then changed its mind.” — An anonymous viewer’s recollection.

From Secret Weapon to “Atomic Tourism”

The 1951 broadcast wasn’t just a news event; it was a pivot point in Cold War propaganda. Before this, the Manhattan Project had been a “ghost.” By bringing the tests to TV, the U.S. government achieved two things:

1. Deterrence: It showed the Soviets that America’s nuclear arsenal was functional and terrifying.
2. Normalization: It attempted to make the bomb feel “contained” and “manageable.” If you could watch it from your sofa while drinking a beer, how scary could it really be?

This led to the bizarre era of “Atomic Tourism,” where Las Vegas hotels advertised “bomb-watching parties” on their balconies, pairing mushroom clouds with martinis.

The 2026 Perspective: Why It Still Haunts Us

Today, we are saturated with “war as content.” We see high-definition drone strikes on TikTok and live-streamed insurgencies. But the 1951 KTLA broadcast was the “Patient Zero” of this phenomenon.

It taught us that humanity has a dark, voyeuristic fascination with its own destruction. As we mark the 75th anniversary, we aren’t just looking back at a technical achievement; we’re looking at the moment we stopped fearing the bomb enough to start watching it for entertainment.

The Legacy of “Operation Ranger”

While that first broadcast was a local L.A. feat, it paved the way for the 1952 nationwide broadcast that reached 35 million people. It turned the Nevada desert into a stage and the American public into a permanent audience for the nuclear age.

The “Atomic Party”: When the Mushroom Cloud Became a Brand

If the 1951 broadcast brought the bomb into the living room, the culture that followed brought it into the kitchen, the bar, and the beauty salon. This “Atomic Chic” era was a psychological defense mechanism—a way for a terrified public to wrap their heads around a force that could end the world by turning it into a theme.

1. The “Atomic Cocktail”: Toasting the Apocalypse

In the wake of the live Nevada broadcasts, Las Vegas became the “Atomic City.” Bartenders at the Sands and the Flamingo competed to create the most explosive drinks.

• The Recipe: Most “Atomic Cocktails” were heavy on vodka, brandy, and champagne, designed to give the drinker a “glow.”
• The Experience: High-end hotels advertised “Sky Rooms” with floor-to-ceiling windows facing the test site. Tourists would sip these potent mixes at 5:00 AM, waiting for the desert horizon to turn white.

2. Miss Atomic Bomb: The Pageant of Power

Perhaps the most surreal byproduct of the era was the “Miss Atomic Bomb” beauty pageant. In a move that feels like dark satire today, the Las Vegas Chamber of Commerce began crowning winners adorned with mushroom cloud props.

• Copa Girl Lee Merlin (1957): The most iconic image of this era features Merlin wearing a fluffy, white cotton-wool cloud attached to her swimsuit.
• The Message: By linking the most destructive weapon ever made with the “all-American girl,” the government and local businesses successfully “domesticated” the apocalypse. It made the nuclear age feel sexy, safe, and distinctly American.

3. Atomic Toys and Uranium Fever

The mania didn’t stop at the bar. It went straight to the playroom.

• The Gilbert U-238 Atomic Energy Lab: Released shortly after the 1951 broadcasts, this was a high-end chemistry set for kids that included real radioactive sources (low-level, but still).
• Pop Music: Songs like “Atomic Baby” and “Uranium Fever” climbed the charts, turning the hunt for radioactive ore into a catchy tune.

4. The 2026 Reflection: Dark Voyeurism

Looking back 75 years later, this “Atomic Party” feels like a fever dream. We see it as a coping mechanism: if you could drink it, dance to it, and dress like it, maybe it wouldn’t kill you.

The 1951 KTLA broadcast was the spark that lit this fuse. It turned a military secret into a pop culture icon. It was the moment we decided that if we had to live in the shadow of the cloud, we might as well have a drink in our hand while we watched.

The Enrollment Cliff Meets the AI Avalanche

Global enrolments are projected to grow through 2035, but the growth is uneven and brutally competitive. Asia and Africa expand, while countries like the United States and parts of Europe face stagnation or outright declines. That means more institutions chasing fewer domestic students in key markets at exactly the moment when alternatives to a traditional degree are exploding.

At the same time, governments and employers are pushing hard on skills, flexibility, and affordability. A major forecasting report warns that AI, digital transformation, and new funding pressures will reshape enrolment patterns, program portfolios, and income models. Put bluntly: there will still be plenty of learners, but they won’t automatically walk through the nearest campus door and sign up for a four‑year degree at any price.​

AI dramatically tilts the playing field. A motivated learner can now combine open online courses, employer micro‑credentials, AI tutors, and simulations into a customized learning journey that looks a lot like a degree—minus the tuition bill. If a university can’t clearly articulate what it adds beyond that stack—network, mentoring, social capital, high‑stakes practice, real assessment—it’s already in trouble.

Robots Took the Back Office. Next Up: The Lecture Hall.

Inside institutions, AI has started where disruption is easiest to sell: the back office. Strategy experts point out that up to a third of current work tasks are automatable, and universities are under pressure to cut costs and “do more with less.” Scheduling, help desks, basic advising, financial‑aid Q&A, recruitment chat, preliminary grading, and IT triage are all either already being handled by AI or on the near‑term roadmap.​

Commentators describe this as AI moving from “tool” to infrastructure, with early‑moving institutions wiring AI into advising, enrollment, learning support, and operations. As that happens, the performance gap between AI‑operationalized universities and laggards widens: faster service, lower overhead, better data, more personalization. Institutions that refuse to automate will look slow, expensive, and weirdly analog within a few years.

The next front is the lecture hall. Analysts expect synthetic or AI‑assisted instructors to become increasingly common, especially in large, lower‑level courses and online programs. AI tools already draft syllabi, generate quizzes, tutor students, run adaptive learning pathways, and even moderate discussion forums. As quality improves, the idea of paying thousands of dollars for a one‑to‑many lecture starts to look less like “higher learning” and more like an outdated content‑delivery service.

Students Are Already Living in the Future

The most dangerous myth on campus is that AI is an “optional” technology students can choose to ignore. Surveys now show extremely high levels of student engagement with AI tools for search, explanation, summarizing readings, and testing out ideas, often without formally involving instructors. For many, AI is simply part of how they think and work.​

Experts warn that higher education’s reluctance to take AI seriously creates a false sense of security. While committees debate policies and pilots, a “shadow system” grows: students using AI to learn, write, and prepare for jobs in ways that bypass institutional structures. If universities do not intentionally design around this reality, they effectively outsource core teaching functions to whatever commercial tools students happen to grab.​

By the time faculty in 2028 realize their course evaluations are really a referendum on how their teaching compares to a free AI tutor, the market will have already shifted. Learners will gravitate toward institutions that treat AI literacy as a core competency, not an honor‑code violation, and that help them use these tools responsibly and effectively.

Only Adaptive Institutions Deserve to Survive

Adapting is not about slapping “AI” into a marketing brochure. It’s about redesigning the core functions of a university around a world where AI is ubiquitous.

1. Curriculum
   Programs that treat AI as a side topic are missing the plot. Policy reports stress that equipping students with AI and digital skills is now a priority for governments and institutions. Every discipline—from nursing to law to engineering to the humanities—will be reshaped by automation, data, and new human–machine partnerships. Surviving institutions will bake AI literacy, data literacy, and ethics into general education and majors alike, rather than ghettoizing them in a single “tech” course.​
2. Pedagogy and Assessment
   When AI can generate a passable essay in seconds, the value of traditional assignments collapses. Analysts call for a rethinking of assessment and a move toward more authentic, project‑based, and performance‑based evaluations. That means in‑class work, oral defenses, labs, studio critiques, long‑term projects with messy constraints—places where human judgment, context, and collaboration are central, and where AI is a tool, not a ghostwriter.​
3. Operations and Cost Structure
   Forecasts for 2035 emphasize that AI and digital transformation will force institutions to change their operating models, not just their tech stack. Automation will allow leaner administrative structures and more targeted student services, but only if leaders are willing to actually cut redundant roles, retrain staff, and redesign workflows. Institutions that cling to 20th‑century bureaucratic structures while competitors deliver faster, cheaper, AI‑enabled services will lose on both cost and experience.​
4. Business Model and Credential Strategy
   As micro‑credentials, employer programs, and alternative providers scale, governments and analysts expect “new forces” to reshape how universities earn income and attract students. Survivors will unbundle degrees into shorter credentials, stackable pathways, and subscription learning. They will partner aggressively with employers, local ecosystems, and global platforms instead of assuming that “the degree” is the only product that matters.​

The AI-First Universities Are Coming

A new class of competitors is emerging: AI‑first universities and AI‑saturated programs that treat automation and personalization as the default. Commentators describe institutions where AI agents help design and deliver learning, run much of the student‑support layer, and orchestrate adaptive coursework at scale. These players pair lean cost structures with high‑touch human roles: master instructors, mentors, and network builders.

Reports argue that by the late 2020s, reduced cost and personalization will create major marketplace advantages for agile institutions. Some colleges will simply close; others will be forced to slash faculty and staff to remain price‑competitive. The winners will be those that move early—integrating AI deeply, defining responsible‑use norms, and communicating a clear value proposition that goes beyond “we also have a chatbot.”

Adapt or Get Out of the Way

The most honest way to read the current research and forecasting is this: higher education is not going away, but a lot of individual institutions will. Growth will happen in systems and organizations that align with an AI‑driven, skills‑hungry, globally competitive world. Others will quietly merge, shrink into niche providers, or close when their combination of cost, speed, and outcomes can no longer be justified.​

AI is not a passing fad, and it is not something universities can cordon off in a single policy document. It is becoming part of the infrastructure of learning, work, and daily life. Institutions that treat it that way—redesigning curriculum, assessment, operations, and credentials around an AI‑saturated reality—have a shot at thriving. Those that do not are effectively choosing to be left behind.

If higher education has a future worth defending, it belongs to the institutions that are willing to adapt in public, at speed, and in partnership with their students and communities. Everyone else is just waiting to be disrupted.

The New Cognitive Division of Labor

For most of history, we offloaded boring mental work to tools: abacuses, calculators, spell‑check, Google. AI is different because it doesn’t just handle arithmetic or search; it volunteers to handle almost anything that feels like effort: drafting, outlining, explaining, brainstorming, even deciding what questions to ask in the first place.​

A recent MIT study on AI‑assisted writing found that people who relied on a model produced good‑looking essays but showed lower cognitive engagement and weaker memory for what they wrote. Researchers called this “cognitive debt”: like financial debt, you get convenience now and pay interest later in the form of shallower understanding and weaker recall. The more you swipe the AI card, the fuzzier your own thinking becomes.

The unsettling part: we are not just outsourcing work; we’re outsourcing the exact parts of thinking that build long‑term skill—wrestling with ambiguity, holding multiple possibilities in mind, feeling the friction of a hard idea.

I Prompt, Therefore I Am?

The easiest way to use AI is as an answer machine: “Tell me what to say.” But some researchers argue that’s the wrong mental model. They suggest treating AI as a “tool for thought” or a Socratic partner that asks questions, surfaces alternatives, and forces you to justify your intuitions.

That sounds noble, but it raises a nasty question: who sets the default? One essay from Oxford’s computing group puts it bluntly—who decides the epistemic and moral priors baked into models that will sit in the middle of universities, courts, clinics, and offices? If your primary thinking surface is a system tuned by a handful of companies and labs, then even when you are “thinking for yourself,” you might be doing it inside someone else’s invisible frame.

You used to inherit your worldview from your family, culture, and media. Now you also inherit it from the gradient descent of a trillion‑token dataset.

Critical Thinking in a World That Auto‑Completes

Big tech and policymakers keep chanting “we need critical thinking,” but the incentives of most AI tools run in the opposite direction. Interfaces are optimized for speed and fluency: you ask, it answers, you move on. Yet surveys of knowledge workers show a consistent worry that generative AI is quietly eroding habits of reflection and scrutiny.

A Microsoft research team reviewing GenAI and critical thinking found a pattern: when AI explanations are presented as polished statements, people accept them too readily; when they’re framed as questions and challenges, people reason more carefully. In other words, UI copy and prompt style can push entire populations toward either “auto‑pilot acceptance” or “active interrogation.”​

That’s a bizarre new design problem: every micro‑choice in how we present model output—confident tone vs tentative, one answer vs a menu of competing hypotheses—nudges the mental posture of millions of users. We’re not just designing tools; we’re tuning the global distribution of doubt.

The White‑Collar Rust Belt

Zoom out to the macro level. One Atlantic piece sketches a worst‑case future where AI doesn’t produce a quick, dramatic robot takeover, but a slow‑motion downgrading of educated workers. Office jobs get atomized into smaller, more automatable tasks. Earnings slide. Unemployment insurance, built for brief dips, buckles under long‑term displacement.

The twist that should scare knowledge workers most is this: the jobs likeliest to erode are precisely the ones that feel safe because they’re “cognitive.” If we train an entire generation to rely on AI for drafting, analysis, and decision support, we might create a white‑collar class that is both economically exposed and mentally deconditioned—people who lost bargaining power and lost the muscles they’d need to reinvent themselves.

That combination—hyper‑unemployment plus hollowed‑out cognitive habits—is more dangerous than either on its own.

Building Friction On Purpose

So what’s the intelligent move if you don’t want to become a prompt‑dependent NPC in your own story?

A few research threads point in the same direction:

• Design “productive struggle” into AI tools: they should ask you to predict before revealing answers, highlight gaps, and force you to choose between conflicting explanations.
• Use AI to amplify curiosity instead of bypassing it: surface surprising counter‑arguments, weird edge cases, things that don’t fit your current model.
• Treat cognitive debt like financial debt: it’s fine to take on some for speed, but track it, pay it down periodically with slower, manual thinking, and don’t live permanently on credit.

One USC Annenberg essay pulls an elegant trick: it lists a whole catalog of AI ethical issues—bias, privacy, opacity, job loss—only to reveal at the end that the entire piece was generated by a model. The author’s point is simple and vicious: if you didn’t question the voice lecturing you about AI ethics, you just experienced the problem in real time.​

That’s the kind of move we probably need more of—systems and stories that force us to feel where we’re giving away our thinking, not just tell us.

The interesting question isn’t “Will AI replace humans?” It’s: Which layers of our own minds are we willing to hand over first—and what happens if we guess wrong about which ones are safe to lose?