Peter Pomerantsev visited Penn State at the end of March, when he was just back from a trip to Ukraine. He joins "Democracy Works" to discuss what he saw there, as well as how American media is covering the war. He also talked about the similarities between Ukraine and the United States when it comes to being vulnerable to Russian disinformation — and how both countries can strengthen democratic media.
AI is unpopular yet widely used. Explore how citizen-led “crackpot schemes” could shape AI policy, protect jobs, strengthen democracy, and maximize AI’s benefits while reducing its risks.
AI is unpopular. And nearly a billion people use ChatGPT.
AI is destroying jobs. And fields predicted to have been eliminated by AI, like radiology, continue to grow and leverage the technology to improve their work.
AI is wrecking the environment. And state officials are learning from hyperscalers how to run the grid more efficiently.
A black or white conception of AI promises to either deny its benefits or exacerbate its shortcomings. Yet, there’s no easy way to ensure a proper balancing of AI’s promises and perils. While every major AI CEO has pledged to "maximize the benefits and mitigate the risks,” they have left out the details of how to do so.
It’s unlikely they or anyone will ever have such a definite plan. But, if we want to avoid repeating what we did with nuclear power — locking a technology away and letting its best uses go unrealized — we need bold ideas from outside the rooms where AI policy currently gets written. We need “crackpot schemes.”
Consider Social Security. Most Americans, asked who we should thank for it, would name FDR. The better answer is a doctor in Long Beach. In 1933, he wrote a letter to the editor proposing a $200 monthly pension for every citizen over 60, funded by a sales tax, with one catch: recipients had to spend the money within thirty days. Two goals, one scheme. Support the elderly. Sustain a Depression-era economy.
The idea caught fire. Clubs formed across the country, 2.2 million members strong. Ten million Americans signed petitions. By 1935, a majority of the country backed it. The plan itself never became law. Critics called it a “crackpot scheme.” But the Social Security system we have today is unimaginable without that op-ed from a doctor with minimal policy chops or political clout but plenty of experience seeing people struggle to achieve financial security in their golden years.
The best policies often come from unexpected places.
Now consider the policy challenges posed by AI. The Bureau of Labor Statistics projects 3.1 percent job growth by 2035. Forecasters at Metaculus project a 1.4 percent loss. Expert forecasters also expect unemployment for recent college graduates to hit 13.5 percent in the near future, with underemployment climbing to nearly 60 percent. Nobody knows for certain how this economic transition will play out. We are entering a volatile decade, and the people most exposed to it are the people with the least say in how it gets governed.
AI must be developed for us and by us if it is going to earn our trust. People support what they help build. As of now, there are few clear mechanisms for people to share their perspective on what they need to feel like AI is a net positive for their wellbeing and the wellbeing of their community. That needs to change.
Here are a few of the “crackpot schemes” I’ve heard about when talking to everyday Americans eager to know that AI progress does not have to come at the expense of Average Joes and Janes.
First, a mandatory national service program with civil and military tracks for Americans between 17 and 19. It would signal that the country values young people's judgment at exactly the moment AI is making judgment more valuable than ever. It would build the interpersonal experience no model can replicate. It would help revive the civic glue that holds this country together.
Second, recurring citizen assemblies at the local, state, and federal levels that provide a representative set of Americans with the chance to share their viewpoints in a calm, deliberative fashion. Polls focused on likely voters may hold sway with political campaigns but do little to help guide actual policymaking. Gathering people in a room to discuss tangible, specific policies and related trade-offs may bolster the ability of politicians to plan for and respond to many different possible futures.
You may regard these as "crackpot schemes" with no odds of passing. Good. The doctor's pension plan was wrong on the numbers, wrong on the mechanism, and wrong on the funding. It was also the most important policy proposal of the decade, because it forced a country that had decided old age was a private problem to admit it was a public one. Your AI proposal does not have to be right. It has to be loud enough to make the wrong answer indefensible. Now share yours.
Kevin Frazier is the Director of the AI Innovation and Law Program at the University of Texas School of Law.
A critique of Steven Rosenbaum's The Future of Truth and the irony of AI-generated errors in a book warning about AI, truth, trust, and democratic responsibility.
A democracy is only as robust and vibrant as the citizens who sustain it. Self-government depends upon people willing to deliberate honestly, reason carefully, and exercise judgment responsibly. With the emergence of AI, this obligation becomes even more consequential because these powerful systems can either deepen human agency or quietly erode it. They can either help citizens think more clearly and participate more meaningfully, or they can encourage the outsourcing of judgment itself and the slow substitution of synthetic plausibility for human responsibility.
Imagine, then, publishing a book warning humanity about the epistemological collapse supposedly ushered in by artificial intelligence. Imagine assembling endorsements from solemn guardians of the humanities, critics of automation, custodians of truth, defenders of interpretation against probabilistic sludge. Imagine presenting yourself as a kind of intellectual fire marshal standing before a burning building, yelling that people must immediately stop playing with matches.
And then imagine your own book turns out to contain fabricated citations, hallucinated references, invented quotations, and synthetic confabulations produced by the very systems whose dangers you claim to expose.
Well, astonishingly, that is precisely what happened with the recently published book The Future of Truth by Steven Rosenbaum.
And how did Mr. Rosenbaum respond when readers discovered that the book warning us about the epistemological dangers of AI had itself smuggled in fabricated citations and synthetic quotations into its pages? With mortification? With embarrassment? With contrition? With the ordinary human shame that arises when one is caught violating one’s own stated principles so completely and so publicly?
Of course not.
Instead, we were treated to a spectacle of post-hoc rationalization masquerading as profundity. The collapse of the book’s credibility was repackaged as further evidence of the book’s importance! The failure itself became the thesis, as it were. As one reads along, one almost expects him to announce that the inaccuracies only proved how urgently we must read his inaccurate book about inaccuracies. And then, defying the ridiculous, that was precisely what Rosenbaum did: He said, “These A.I. errors do not, in fact, diminish the larger questions that the book raises about truth, trust and A.I.”
Well, of course, they do not diminish them, but they do diminish the book and its author spectacularly.
The book is about truth and trust, and yet the author himself failed at the elementary practice of verification. How can one deliver a sermon on epistemic rigor while outsourcing one’s footnotes to a stochastic autocomplete engine and then declare oneself vindicated when people notice?
Even worse is this peculiar modern move to hide behind a machine as though it were an unruly intern who slipped something embarrassing into the manuscript. “It’s the AI that made the errors, not me” is rapidly becoming the twenty-first-century version of “mistakes were made.”
No, Mr. Rosenbaum: You made the errors.
You used the tool. You signed the contract. You accepted the authority of authorship. You placed your human name on the cover rather than listing the author as “Large Language Model, Version Unknown.”
The entire point of being human is that we are not merely output systems. We are capable of judgment, deliberation, doubt, scruple — that’s our human bailiwick. We can stop, we can verify, we can say: Hey, this sentence sounds plausible, but I should check whether it corresponds to reality before presenting it to thousands of readers. That is our slice of the divided labor pie.
And then, when we fail, we, humans, are supposed to be capable of something machines are not: Contrition. And actual contrition. Not sleek, too-clever- (or maybe too-stupid-) by-half post-hoc rationalization disguised as profundity; nor should we be converting our embarrassment into an advertisement for the very thesis whose collapse occasioned the embarrassment; nor should we indulge ourselves in insulting the intelligence of our readers by pretending that a humiliating failure of scholarship somehow deepens our work’s philosophical importance.
There is something oddly revealing in all this. Many critics of AI speak as though the machine itself were the central moral actor, as though silicon were secretly plotting the liquidation of truth. But the more immediate spectacle is much pettier and much more human: Vanity amplified by convenience, and obviously a whole lot of good old laziness. People lusting for authority of thought but putting in the actual work of thought. Wanting the appearance of mastery while quietly delegating the difficult work of verification to a predictive engine, and then blaming it for not doing a perfect job.
In any case, at the very least, the irony is exquisite. A book warning us about epistemological collapse becomes itself an artifact of epistemological collapse. A denunciation of synthetic authority turns out to depend upon synthetic authority. The critic becomes the demonstration. The cautionary tale writes itself. All of which is amusing enough, as these things go, but when the matter concerns the quality of democratic participation, we need to take these questions far more seriously than this.
Ahmed Bouzid is the co-founder of The True Representation Movement.
As someone who grew up with the internet, it has been hard to set boundaries around screen time. I grab my phone instinctively—waiting for my coffee order or in a crowded elevator—as one or two minutes of stillness is apparently impossible. And that in itself concerns me.
It’s when I’m winding down to sleep, telling myself I can open TikTok or Instagram (my go-tos) for a bit, and suddenly two hours pass by, that I admit I’m practicing a harmful habit. This is “doomscrolling,” spending excessive time scrolling through social media content, typically negative news, that usually makes the person feel anxious, sad, or angry.
Nearly 1 in 3 Americans (31%) who use social media said they doomscroll, with younger generations at the forefront of this behavior (46% of millennials and 53% of Gen Z respondents), according to a 2024 Morning Consult survey.
Although everyone is susceptible to doomscrolling, according to Harvard Medical School Professor Dr. Richard Mollica, women and people with a history of trauma are also particularly vulnerable to doomscrolling.
While “doomscrolling” at night does not seem to affect everyone’s sleep the same way, significant research shows that it can greatly impact people’s emotional and physical health.
The repetitive consumption of negative content can “activate the body’s stress response,” which leads to elevated levels of anxiety and other mental health conditions. In the past decade, studies have linked the rise of social media use and doomscrolling to an increase in anxiety among young adults across the country.
Digital screens also emit blue light, which can delay the secretion of melatonin (a natural hormone that is important for sleep). Aside from poor sleep quality, blue wavelengths are also associated with other health problems, including mental health issues and increased risk of some cancers.
Charge your device across the room at night. Move that phone away from you. Give yourself a few minutes before bed (put a timer if needed) to check or respond to any messages. Then charge your phone, iPad, etc. across the room for the evening. This prevents instinctively reaching for your device. (And if you have an alarm for the morning, that’s even better—now you have to get up in the morning to turn it off.)
Instead, keep something like a book, comic, or journal by your bed if you can’t sleep. Or play an audiobook, podcast, or white noise from across the room to sleep.
Limit your screen time during the day. Be intentional about your social media time, and set a timer for how long you can scroll. Find times in your routine when it’s appropriate to use your phone, but you’re not tempted or able to spend more than 10-15 minutes on it.
You can also set daily limits on certain apps on your devices. For Apple products, you can use the “Screen Time” feature in settings. On Android, you can turn on the “Digital Wellbeing” feature. There are also separate apps that do this—I personally enjoy “Focus Friend” by Hank Green; I use it daily.
And one in doubt, temporarily (or permanently) delete an app from your phone. You can also hide an app from your home screen so you’re less inclined to instinctively click on it.
Read your local paper if you have one! Newspapers are a timeless way to stay up to date. To me, there’s something special about the feel of a new issue with a cup of coffee in the morning or tea before bed. When you grab your local paper, you’re simultaneously engaging with and supporting your immediate community—something we need more of.
Long-form digital content. When it comes to digital media, there are plenty of ways to effectively inform yourself on current events—and this is through longer form content.
Use social media as a starting point to discover timely events or issues that interest you. Make a note of these topics, and search for in-depth content that focuses on them: YouTube video essays, documentaries, podcasts, Substacks, etc.
In a time of endless digital content and large social media presences, there is a lack of genuine connection between people.
A few times a week, my father and I call each other about some event in the news or an article that we’ve read, and we discuss it. There is something more meaningful and cathartic when you can talk through current events with the people around you.
Although political conversations have become so taboo in recent years, the only way we can have productive, open-minded conversations is through gentle practice.
If I have the energy to engage in a debate and have some of my ideas challenged, I bring up certain topics with my cousins (we are all really close, politically diverse, and we know when to pause the conversation if needed). If I feel emotionally exhausted and wish my feelings to be validated, I have other relatives and friends to reach out to. I am incredibly privileged to have diverse opinions and voices around me, because being both challenged and comforted holds incredible value. Mindlessly scrolling through my social media feed does neither.
Basically, doomscrolling is never productive. We have the right and ability to inform, empower, and care for ourselves and our communities. We just need to get off our phones.
Belen Dumont is the Associated Editor of The Fulcrum.
Technology is currently transforming the economy faster than established institutions and societal choices can keep pace. Decisions made now regarding ethics, training, and governance will ultimately determine who benefits from these advancements and "who gets pushed aside".
In a recent episode of The Fulcrum Roundtable, Kevin Frazier, a collaborator with The Fulcrum, discussed two of his articles: Build Better AI and Layoff Headlines Keep Coming, Policy Answers Don't. Here’s One Solution.
The senior fellow at the Abundance Institute and director of the AI Innovation and Law Program at the University of Texas School of Law—emphasized that the future of responsible artificial intelligence requires moving beyond corporate mission statements toward active public participation and localized policy solutions.
- YouTube youtu.be
Frazier noted that while major AI labs like Google, Meta, Anthropic, and OpenAI often pledge to "harness the positive attributes" of AI while "mitigating the negative," such aspirational statements are often common sense—or "duh"—principles that lack specificity. He argued that society must be far more specific about identifying concrete "red lines" and specific positive use cases to prevent a purely reactive posture. Building responsible AI "demands far more than aspirational mission statements".
To achieve this, Frazier recommended a three-pronged approach for citizens to help shape the path forward. First, individuals should actively experiment with AI tools for "pro-social" tasks—such as article translation or researching diverse perspectives on voting issues—to understand their capabilities. Second, users should transparently "share when those use cases go right and share when those use cases go wrong" to help popularize tools that align with community values. Finally, Frazier urged political engagement, noting that since late 2022, over 1,700 AI-related bills have been introduced at the state level. He encouraged citizens to contact state representatives and city council members to help shape early-stage policy.
The transition to an AI-driven economy presents significant risks of "rocky" transition periods. While economists predict aggregate long-term job growth, the timing, location, and nature of these new roles remain uncertain, potentially leaving entire communities behind. Frazier criticized current retraining efforts, such as the Workforce Innovation and Opportunity Act (WIOA), as being underfunded, fragmented, and lacking coordination. He specifically highlighted poor metrics that prioritize the speed of re-employment over job quality, which can push laid-off workers into new roles that remain highly vulnerable to AI displacement. He proposed "transition launchpads"—local hubs where displaced workers can receive training based on the specific demands of local employers.
Frazier also warned of an emerging "AI abyss" that could widen existing digital divides. He identified several groups at risk: those in rural and urban poor communities without affordable high-speed broadband; older Americans, as recent polling indicates a plurality of Boomers and the Silent Generation have not meaningfully engaged with AI; and non-English speakers, as many AI models "may not be as capable" when prompted in other languages.
Drawing a historical parallel to the 1930s Rural Electrification Administration and its "electrical circus," Frazier suggested using trusted local messengers—such as public librarians, teachers, and civic leaders—to demonstrate responsible AI use and protect community members from scams.
I invite you to spend some time reading his articles by clicking HERE.
Hugo Balta is the executive editor of The Fulcrum and the publisher of the Latino News Network.
Combating disinformation at home and abroad
Trump is stuck between two realities. Neither serves the American people