In an era of increasing social isolation, platforms like Facebook and Twitter are among the most important tools we have to understand each other. However, social media can also lead to polarization and misunderstanding as author Chris Bail discusses in this episode of Democracy Works.
For many people, the current anxiety about artificial intelligence feels overblown. They say, “We’ve been here before.” Every generation has its technological scare story. In the early days of automation, factories threatened jobs. Television was supposed to rot our brains. The internet was going to end serious thinking. Kurt Vonnegut’s Player Piano, published in 1952, imagined a world run by machines and technocrats, leaving ordinary humans purposeless and sidelined. We survived all of that.
So when people today warn that AI is different — that it poses risks to democracy, work, truth, our ability to make informed and independent choices — it’s reasonable to ask: Why should I care?
The answer is not that AI is evil, or has its own agenda, or is plotting against us. The real concern is quieter and more subtle. AI changes how judgments are formed, and it does so at a scale and speed no previous technology has reached.
Most fears about AI focus on dramatic outcomes: mass unemployment, autonomous weapons, runaway systems. Those are serious issues, but they miss a more immediate shift already underway.
AI doesn’t replace human judgment by force. It replaces it by being effective, fast, confident, and coherent.
To see why, it helps to look at how we’ve traditionally answered questions.
Years ago, if you wanted to understand a topic, you went to a library. You found a few relevant books, read them carefully, compared arguments, evaluated sources, and slowly formed your own understanding. The process was time-consuming, but judgment — deciding what mattered and what made sense — was entirely yours.
Then computers and search engines arrived. With Google, you could instantly find thousands of related sources. But you still had to choose which ones to read, decide what was credible, assemble the facts, and form your own conclusions. The machine found information, but humans still did the thinking.
Now enter AI.
You ask a question once. The system reviews thousands — or millions — of sources, past discussions, data points, and arguments, and then almost instantly delivers a coherent answer that often makes remarkable sense. You can accept it, refine it, or go deeper, but the first draft of your understanding — your judgment — is no longer yours. It arrives already assembled.
Judgment isn’t just opinion or instinct. It’s the cognitive process of weighing evidence, context, experience, uncertainty, and consequences in order to decide what to believe or what to do.
Judgment is demanding. It requires grappling with uncertainty, weighing tradeoffs, and accepting responsibility. It’s also where accountability lives: when things go wrong, some person or body owns the decision.
AI doesn’t eliminate this process. It changes when and how it happens.
Instead of forming judgments from raw material, humans increasingly review judgments that have already been synthesized — often fluently, confidently, and persuasively.
That shift matters.
AI systems feel “smart” not because they think like humans, but because they can absorb and organize enormous amounts of information — far more than any individual could manage — and present it clearly and quickly.
Ask about a legal issue, and the system draws on statutes, cases, patterns, and prior interpretations. Ask about policy, and it references history, comparative examples, and arguments on all sides. The result is often impressively accurate, or at least plausibly so.
There’s nothing mystical about this. But the effect on humans is powerful.
When an answer is:
…it feels authoritative. It reduces uncertainty. It saves effort. It saves time.
And that creates a quiet temptation: Why wouldn’t I trust this?
What makes this shift especially powerful is that AI can be engaged the way we engage other people.
Search engines require you to think like a machine — keywords, fragments, trial and error. AI lets you think like yourself. Like you're talking to a friend. You can explain your situation, add nuance, express uncertainty, introduce complications, or explore alternate ideas. The system understands what you’re saying, no matter how complex, and follows the thread.
For the first time, a machine doesn’t just retrieve information — it participates in the process of reasoning. It talks back, suggests, and discusses with you.
That matters because humans trust explanations that feel conversational and attentive. We trust fluent explanations. We are influenced by tone, coherence, and responsiveness. When something engages us conversationally, it doesn’t feel like an external tool. It feels like an extension of our own thinking or input from a friend.
At that point, the line between assistance and influence becomes thin.
This is not a sudden takeover. It’s a progression.
First, AI provides information.
Then it provides analysis.
Then it offers recommendations.
Eventually, people defer — not because they are forced to, but because ignoring the system starts to feel inefficient or even irresponsible.
Judgment doesn’t disappear. It atrophies.
And when something goes wrong, accountability becomes blurred:
That’s not tyranny. It’s drift. No one takes control from us; we slowly hand it over because it works.
A single biased person can influence dozens.
A biased book can influence thousands.
An AI system can influence millions — instantly, continuously, and adaptively.
Now imagine not one AI system, but thousands. Built by different companies, trained by different people, optimized for different goals — profit, engagement, persuasion, efficiency, ideology.
Some will be careful and transparent. Some will not. Some will be neutral. Some will be highly motivated.
Two people could ask the same question and receive different facts, different framing, and different conclusions — all delivered with confidence.
At that point, disagreement isn’t just about opinion. It’s about perceived reality itself — what facts people think exist and which ones they trust.
Societies and individuals are stressed under those conditions.
We must also understand that AI doesn’t respect borders. A system trained in one country can shape opinions, decisions, markets, and political discourse far beyond where it was built.
For other global risks — nuclear weapons, climate change, pandemics — we eventually recognized a basic truth: if the risk is global, oversight must be global.
While our efforts with these global risks have been imperfect, the essential international coordination that allows institutions that monitor risks, share information, and establish norms has caused us to pause, reflect, and develop strategies, guidance, and cooperative efforts.
With AI, that coordination barely exists.
Private companies largely mark their own homework. Governments are regulating unevenly. And a competitive race dynamic discourages restraint, transparency, and caution. No country wants to fall behind. No company wants to slow down. The result is fragmentation at exactly the moment shared standards matter most.
This is especially dangerous for open societies. Democracies depend on shared facts, public trust, and slow, deliberative institutions — precisely the things AI systems place under the greatest pressure. When judgment is accelerated, personalized, and scaled globally without common rules, the foundations of democratic consent begin to erode.
Markets are excellent at optimizing for speed, convenience, and profit. They are not designed to optimize for shared truth, civic trust, or democratic stability.
Left entirely to competition, the incentive is clear: build systems that keep people engaged, tell them what they want to hear, and do it convincingly.
That doesn’t require bad intentions. It only requires misaligned incentives.
History shows that when powerful technologies reshape how people understand the world, coordination matters.
Not because AI will suddenly take control.
Not because humans will become irrelevant.
But because we may be slowly delegating the first draft of our understanding to systems we did not collectively design, govern, or agree upon.
AI isn’t dangerous because it’s evil.
It’s dangerous because it’s so convenient and fast.
And once we notice how much judgment we’ve handed over, it may already be beyond our control.
And all of this is unfolding not within shared global rules or institutions, but through a fragmented international race involving governments, companies, and systems that operate far beyond any single country’s control.
Jeff Dauphin is currently retired - Blogging on the "Underpinnings of a Broken Government." Founded and ran two environmental information & newsletter businesses for 36 years. Facilitated enactment of major environmental legislation in Michigan in the 70s. Community planning and engineering. BSCE Michigan Technological University.
AI-generated “nudification” is no longer a distant threat—it’s harming students now. As deepfake pornography spreads in schools nationwide, educators are left to confront a growing crisis that outpaces laws, platforms, and parental awareness.
While public outrage flares when AI tools like Elon Musk’s Grok generate sexualized images of adults on X—often without consent—schools have been dealing with this harm for years. For school-aged children, AI-generated “nudification” is not a future threat or an abstract tech concern; it is already shaping their daily lives.
Last month, that reality became impossible to ignore in Lafourche Parish, Louisiana. A father sued the school district after several middle school boys circulated AI-generated pornographic images of eight female classmates, including his 13-year-old daughter. When the girl confronted one of the boys and punched him on a school bus, she was expelled. The boy who helped create and spread the images faced no formal consequences.
The case ignited debate over internet safety, deepfake pornography, and school discipline. But it also exposed a deeper truth we are reluctant to confront: decisions made by powerful tech leaders are reshaping childhood faster than schools, parents, or laws can respond—and schools are being left to manage the fallout without the tools they need.
Recent survey data confirms this is not an isolated incident. Researchers found that AI “nudification” is increasingly common in schools, used to harass, humiliate, and exert power over peers. What adults may still perceive as shocking misconduct has, for many students, become disturbingly normalized.
In nearly all 50 states and Washington, D.C., creating and distributing child sexual abuse material is a crime. AI-generated deepfakes, however, present a unique challenge. These images are easy to create, can be shared widely in seconds, and often disappear from platforms just as quickly. Even when perpetrators are identified, the speed, volume, and anonymity of digital sharing make enforcement extraordinarily difficult.
Expecting the legal system to track and prosecute every child and teenager contributing to this epidemic is neither realistic nor effective. If we focus only on punishment after harm occurs, we will always be too late. The goal must be prevention.
Research shows that 31 percent of young people are familiar with deepfake nudes, and one in eight knows someone who has been victimized by them. Girls account for 99 percent of the victims. One in 17 youth and young adults has been directly targeted by AI-generated sexual images—roughly one student in every middle school classroom in the United States. This is not a fringe issue or a moral panic. It is a widespread form of sexual harassment enabled by technology that outpaces our safeguards.
Students need clear guidance to navigate a digital world where a single harmless photo can be transformed into a weapon—sometimes without malicious intent, but with devastating consequences. Yet only 28 states and the District of Columbia require sex education, and just 12 include instruction on consent. This gap has created ideal conditions for the deepfake crisis to flourish.
Without education on bodily autonomy, digital boundaries, consent, and meaningful safeguards from tech companies, young people are left unequipped to recognize the harm in creating and sharing explicit AI images. They are even less prepared to respond when they or their peers become targets.
As a mother, I resist the urge to say simply that parents need to talk to their kids. Parents are essential, but many lack the technical knowledge, consistent access, or awareness needed to explain how these images are created, how quickly they spread, and the profound psychological harm they cause. That is where schools must step in.
As a former middle school teacher, I have sat across from parents explaining the seriousness of emerging online trends long before they reached Facebook groups, GroupMe chats, or parent blogs. Schools are often the first places where this harm appears—and they are uniquely positioned to respond.
Schools can and should provide structured, age-appropriate education that reaches all students, ensures consistent messaging, and creates space for honest discussion. Lessons should include:
Educators already manage cyberbullying, hunger, school violence, and adolescent mental health. Some may ask whether this is one burden too many. But integrating education about AI-generated pornography is not an added responsibility—it is a necessary evolution of student safety in a digital age.
Unlike many victims, both the woman targeted on X and the 13-year-old girl in Lafourche Parish reported their abuse. But for every report, how many students suffer in silence—ashamed, afraid, or unsure whether adults will take them seriously?
While platforms like X attempt to normalize or minimize the harm of deepfake nudification, educators must push back against the idea that this behavior is accessible, acceptable, or consequence-free. That message does not stay online. It reaches classrooms, school buses, and lunch tables. When perpetrators face little accountability and victims are punished for reacting, the lesson students learn is devastatingly clear.
If tech leaders will not fully account for the damage their products enable, schools must act—not through harsher punishment, but through education. Teaching AI literacy, consent, and respect is our strongest defense against a problem that is only growing. Prevention, not discipline, is how we protect children—and how we ensure no more students have to fight back just to be heard.
Julienne Louis-Anderson is a former educator, curriculum writer, and educational equity advocate. She is also a Public Voices Fellow of The OpEd Project in Partnership with the National Black Child Development Institute
Even with the full protection of the First Amendment, the free press in America is at risk. When a president works tirelessly to silence journalists, the question becomes unavoidable: What truth is he trying to keep the country from seeing? What is he covering up or trying to hide?
Democracies rarely fall in a single moment; they erode through a thousand small silences that go unchallenged. When citizens can no longer see or hear the truth — or when leaders manipulate what the public is allowed to know — the foundation of self‑government begins to crack long before the structure falls. When truth becomes negotiable, democracy becomes vulnerable — not because citizens stop caring, but because they stop receiving the information they need to act.
The framers wrote the First Amendment because they understood the danger of leaders who try to control information. They had lived under a government that censored newspapers, punished criticism, and used fear to silence dissent. So they wrote a simple, absolute command: Congress shall make no law abridging the freedom of the press. They were not protecting journalists as a profession; they were protecting the people’s right to know what their government is doing.
Senator John McCain, who died in 2018 after more than three decades of service in Congress, was widely regarded as an influential figure within his party and a leader who valued bipartisan cooperation. Though he and then‑Senator Joe Biden belonged to different political parties, they worked together for years on national security, foreign policy, and veterans’ issues — a reminder that strong leaders can debate fiercely while still upholding democratic norms.
McCain had been warning about the dangers of undermining the press for years. In a 2017 interview on NBC’s Meet the Press (McCain Interview), he said that he “hated the press,” but defended it because democracy cannot survive without it. He cautioned that suppressing a free press is “how dictators get started,” emphasizing that an adversarial press — one willing to challenge leaders and expose uncomfortable truths — is essential to preserving individual liberties.
Yet today, we are watching the very scenario the framers feared. Leaders who dislike scrutiny increasingly turn to a familiar strategy: discredit the press so the public will distrust the facts. They label critical reporting as “fake,” call journalists names, question their motives, and attack their credibility. These tactics are not spontaneous; they are strategic. They are designed to weaken the institutions that hold power accountable and to convince citizens that only the leader can be trusted.
Most citizens only see the tip of the iceberg — the public statements, the press conferences, the televised moments. But journalists uncover what lies beneath: documents, decisions, contradictions, and consequences. When leaders attack the press, they are not reacting to the tip of the iceberg; they are trying to keep the public from seeing what lies below the surface.
And yet, the free press remains one of the most resilient institutions in American life. It has survived the Sedition Act, McCarthyism, and Watergate. Even under relentless pressure, journalists persist in uncovering facts that powerful people would prefer to keep hidden. That resilience is not accidental — it is the design the framers intended.
I once overheard a senior citizen speaking with a friend about the documents found at Mar‑a‑Lago. I don’t remember the exact words, but I remember the look on her face when she said, “They did not tell us that on our news.” It was a moment of quiet realization — the kind that reveals how easily citizens can be kept in the dark without ever knowing it. She wasn’t choosing silence; she was receiving only fragments of the truth.
A friend once asked me, “Why are people so upset? Why don’t they leave President Trump alone? After all, the FBI delivered the documents to him at Mar‑a‑Lago.” I had no words. I simply asked who told her that, and suggested she look deeper to find the truth. Her comment wasn’t malicious — it was the product of a news environment where people can be absolutely certain they are informed while receiving information that is incomplete or inaccurate.
This is why democracies do not collapse in silence. Citizens hear what is happening. They see what is happening. The danger is not that the public is unaware — it is that leaders work to distort, deny, or drown out what the public witnesses. When noise replaces information and denial replaces accountability, leaders can act without scrutiny. A weakened press does not just limit what citizens know — it expands what leaders can hide. Recent events continue to show how easily truth can be distorted or denied. Reporting from the Associated Press (APNews) and Reuters (Reuters.com) on Venezuela shows why independent journalism matters: without it, the public would have no window into decisions that affect national policy and global stability.
Misinformation now spreads at the speed of a click, while truth moves at the speed of verification. Interviewers must be prepared, ask evidence‑based questions, and challenge falsehoods in real time because the public is watching in real time.
Many Americans now rely on one‑sided news sources that reinforce what they already believe. These echo chambers create insulated audiences who rarely encounter conflicting information. The constant barrage of conflict and contradiction becomes emotionally exhausting, making disengagement feel like self‑protection. But disengagement creates a vacuum — and power always fills a vacuum.
Defending a free press does not mean pretending that all news outlets uphold the same standards. Some blur the line between fact and opinion or amplify misinformation. But the failures of some outlets do not diminish the constitutional role of a free and independent press. A democracy cannot function when citizens are misled — nor can it survive when journalists are silenced.
Leaders have a responsibility not only to govern, but to model the democratic values they swore to uphold. That begins with telling the truth — even when the truth is inconvenient or politically costly. The best rule for any leader is simple: if you do not want the public to know about it, then do not do it. Public office is not a private enterprise. Decisions made in the dark almost always betray the people who must live with the consequences.
My mother used to say, “What you do in the dark always comes out in the light.” It is a truth echoed across history, philosophy, and even scripture — Luke 12:2–3 (NIV) and Mark 4:22 (NIV) — the idea that hidden actions eventually become known.
If leaders cannot tell the truth, they should at least refuse to lie. If they cannot answer a question, they should not attack the person asking it. If they cannot be transparent, they should simply say, “No comment at this time,” rather than distort, dehumanize, or deny. Leaders honor the First Amendment and must commit to transparency, correct misinformation rather than amplify it, and respect the independence of institutions that exist to protect the public — not the powerful.
Protecting a democracy built on truth requires action from every institution and every citizen. The press must continue to investigate and publish information about government actions, even when those findings are uncomfortable or politically inconvenient. Journalists must hold leaders accountable, challenge attempts to block or suppress reporting, and uphold the highest ethical standards. Their role as watchdogs is not ceremonial — it is constitutional. They must defend their right to gather accurate information, use the courts when necessary to challenge unlawful restrictions, and explain to the public why a free press is essential to self‑government. A democracy cannot function when journalists are intimidated, sued, retaliated against, or threatened for doing their jobs. Those are the tactics of weak leaders, not democratic ones.
Citizens also carry responsibility. Democracy depends on an informed and engaged public — people who attend meetings, ask questions, write emails, make calls, and speak directly to officials when something is wrong. Citizens must research information to validate facts, support fact‑based journalism, and challenge misinformation respectfully but firmly. They must choose their news sources carefully, avoiding echo chambers that reinforce only what they already believe. They must exercise their rights to speak, record, photograph, and assemble peacefully in public forums — especially when those rights are challenged, as when ICE officers recently tried to prevent protesters from filming. And when necessary, citizens must protest peacefully, as seen in the Jimmy Kimmel case, to defend free expression and the public’s right to know.
Elected officials must model the democratic values they swore to uphold. Congress must stop covering up, repeating, or defending lies. They must stop twisting the truth to protect political allies and stop hiding behind excuses like “I haven’t read it” or “I didn’t know.” Transparency is not optional; it is the foundation of public trust. Leaders must give honest answers — or, when necessary, simply say “No comment” rather than mislead the public. Public office is not a private enterprise. Decisions made in the dark almost always betray the people who must live with the consequences.
A president must respect the First Amendment and the public’s right to know. That means telling the truth — not lying, covering up, or manipulating information. A president must stop attacking, discrediting, or dehumanizing the press, and must stop calling factual reporting “fake” simply because it is unfavorable. If a president does not want an action reported, the solution is simple: do not engage in conduct that cannot withstand public scrutiny. No president should intimidate journalists, instill fear, retaliate against them, or use lawsuits and legal threats to silence reporting — because that is what weak leaders do.
Truth is a public good — and when it is weakened, the entire civic ecosystem suffers. A nation that cannot see or hear the truth cannot remain a democracy. But a people who insist on truth — who demand transparency, challenge misinformation, and hold leaders accountable — will always have the power to protect it, and to pass it forward.
Carolyn Goode is a retired educational leader and national advocate for ethical leadership, government accountability, and civic renewal. She writes about democratic norms, constitutional responsibility, and the role of truth in public life.
As AI reshapes the labor market, workers must think like entrepreneurs. Explore skills gaps, apprenticeships, and policy reforms shaping the future of work.
What do a recent grad, a disenchanted employee, and a parent returning to the workforce all have in common? They’re each trying to determine which skills are in demand and how they can convince employers that they are competent in those fields. This is easier said than done.
Recent grads point to transcripts lined with As to persuade firms that they can add value. Firms, well aware of grade inflation, may scoff.
Disenchanted employees need to spend time training for the jobs of the future, perhaps by working toward a badge or credential. Firms are rightfully skeptical of those, too. After all, there are hundreds of thousands of certificates these days—it’s unclear which are meaningful.
Parents try to convince firms that they remain as skillful as ever by highlighting their earlier work. Here, again, firms may have some questions. A hiring manager may not shake the nagging feeling that extended time off the job may have caused version skills to atrophy.
In a healthy, efficient labor market, it’d be possible for these workers to signal their skills and to find firms demanding such work. The aforementioned barriers all stand in the way of such a market. The introduction of AI makes this labor matching even more difficult. Firms don’t know which skills to seek out because it’s unclear what work will be completed by humans, human-AI teams, or just AI. Workers, too, are at a loss—hoping that the skills they seek to gain align with those demanded by firms over the long run.
In this market failure—when informational asymmetries prevent workers and firms from finding one another in as cheap and timely a manner as possible—it’s tempting to call on the government to step in. The thinking goes that the government can predict which skills will define the future and can set up programs for upskilling and retraining. This logic falls by looking to the advice of many a government officials to lean into computer science. While some firms may demand some individuals with such skills, early returns from the Age of AI suggest demand is dropping.
What’s a student to do? How can someone finally leave their firm and find a better role? How can a mom or dad get back into the office and stay there?
The answer is simple and, perhaps, daunting: we’re all entrepreneurs now. We all must be attentive to market trends, adaptive to meaningful shifts in labor demand, and willing to work in novel and, at times, unpredictable environments. In short, the career ladder may be broken, but it’s been replaced by a career flywheel—studying when necessary, shadowing as a trainee, and working in flexible arrangements.
No one—including AI experts and government economists—can detail the specific set of skills that will result in high-paying work that supports families and sustains the American Dream. Everyone must be willing to take risks—diversifying, deepening, and shifting their skill sets to be as valuable in the labor market as possible.
Politicians ought not try to forecast those skills but rather sustain an entrepreneurial approach to skill development. Three proposals can help channel the necessary entrepreneurial energy without being too prescriptive.
First, emulate South Carolina’s success by encouraging firms to offer more apprenticeships. South Carolina has quietly built one of the most effective apprenticeship ecosystems in the country. Through its registered apprenticeship initiative, the state helps firms offset training costs, coordinate curriculum with community colleges, and design programs that respond to real production needs rather than abstract projections. The result is a pipeline that places people into paid roles while they learn—reducing the risk for workers and giving firms a chance to assess talent in real time.
Second, encourage the creation of skills-based evaluations by high schools and higher education institutions. Workers can better market their services—and firms can hire with more confidence—when skills are legible, portable, and comparable. Today’s degrees obscure more than they reveal. Grade inflation compresses distinctions, transcripts say little about applied competence, and employers are left guessing.
The Department of Education—and state equivalents—can help by issuing guidance that promotes competency-based transcripts, standardized skill taxonomies, and verified portfolios that document what students can actually do. It can also serve as an information clearinghouse, publishing data on which institutions and programs reliably produce particular skills and outcomes.
Third, reform—or abandon—New Deal employment laws with rigid classifications that hinder the ability of workers and firms alike to engage in creative, flexible, and mutually beneficial arrangements. The Fair Labor Standards Act, built for an economy of factory floors and fixed schedules, struggles to accommodate project-based work, part-time experimentation, and hybrid human-AI roles.
Its binary distinctions between employee and contractor, for one, discourage firms from offering flexible pathways and push workers into all-or-nothing choices. Modernizing these rules—by allowing more fractional work, clearer safe harbors, and updated definitions of hours and supervision—would expand opportunity without sacrificing baseline protections.
The future of work will not be handed down in a syllabus or codified in a regulation. It will be discovered—through trial, error, and adaptation—by people willing to build, learn, and pivot. Policy should not pretend to know which skills will win. It should instead clear the runway so more Americans can take off, test their wings, and land somewhere better.
Kevin Frazier is a Senior Fellow at the Abundance Institute, directs the AI Innovation and Law Program at the University of Texas School of Law, and is an Affiliated Research Fellow at the Cato Institute.
