The Golden Moment
Gary Rivlin and the Human Cost of Innovation
The Reporter in the Machine
Gary Rivlin, who resides in New York City, utilizes AI daily, sometimes as often as 20 times a day, particularly when he is writing. He employs AI as an editor, an idea generator, and a research assistant for his work, including the writing of his book, AI Valley.
Gary Rivlin joined us at the Saratoga Book Festival. We then had another 90 minutes with him in Zoom. To generate this story, we uploaded the transcript to NotebookLM and processed it there and in ChatGPT. We call our suite of AI tools the Smartacus Neural Net.
During his creative process, Rivlin may prompt his digital assistant—such as Claude, his preferred model—to synthesize large volumes of interviews or to offer edit suggestions on a paragraph he dislikes. When wrestling with writer's block, he might ask the AI to rewrite a difficult paragraph in five different ways, often finding a "good idea" or the right word he had not previously considered. He emphasizes that he never simply cuts and pastes the output; instead, he uses the AI’s input as a starting point to rework the material and put it in his own voice.
Rivlin maintains that the human must remain the creator and the driver in this partnership, viewing AI strictly as a tool or "co-pilot". He stresses that he must stand by everything published under his name. He reminds himself that, should the AI hallucinate or make a factual error, "it’s Gary Rivlin’s mistake".
That simple declaration encapsulates Rivlin’s career: the human must remain in charge. He calls this moment a “golden moment”—a fleeting window while artificial intelligence is “powerful but not all-powerful.”
“My lament is, I wish we were wrestling with this,” he says. “These models are good, but they’re not nearly as good as they’re going to be. We should be having these basic debates.”
For Rivlin, the debates he means aren’t abstract. They are about surveillance and sentencing, hiring and education, mental health and meaning—the everyday places where algorithms are already making choices once reserved for humans. In AI Valley, his newest book, he returns to Silicon Valley not as a cheerleader or doomsayer, but as a historian of our moral crossroads.
Rivlin’s instinct to interrogate power long predates the age of algorithms. He began in Chicago, where politics was theater and journalism a contact sport. In his twenties, at the Chicago Reader, he launched a column on City Hall and embedded himself in the storm surrounding Harold Washington, the city’s first Black mayor. Those dispatches became Fire on the Prairie (1992), a chronicle of race, reform, and resistance that won the Carl Sandburg Award.
The book revealed what would become Rivlin’s lifelong method: follow the power, trace the money, measure the human cost.
Oakland: The Human Toll
After moving west, Rivlin joined the East Bay Express and immersed himself in a story that would shape his understanding of American inequality for decades to come. The case was devastatingly simple: a drive-by shooting in East Oakland had left 13-year-old Kevin Reed dead. What began as a breaking-news item grew into a years-long inquiry into how violence incubates in communities already ravaged by poverty, drugs, and disinvestment.
Rivlin sat with grieving parents, teachers, and police officers. He walked the blocks where the shooting occurred, tracing how retaliation spiraled into more tragedy. The story became Drive-By (1995), a haunting book about grief, anger, and survival in one Oakland neighborhood. “I wanted to examine the human side of this country’s youth-violence epidemic,” he said—an epidemic that, in his view, was less about gangs than about policy and neglect.
The experience taught him that the forces shaping human suffering are often systemic, invisible, and normalized. A lack of opportunity, he saw, could be just as lethal as a bullet. That insight became the foundation of all his later work. Whether writing about urban violence, predatory lending, or the latest tech boom, Rivlin sought the same recurring pattern: systems that reward a few and wound the rest.
The Dot-Com Crash: Lessons in Hype
In 1999, Rivlin dissected the legal case against Microsoft.
By the late 1990s, Rivlin had turned his eye to Silicon Valley. What fascinated him wasn’t the circuitry—it was the speculation. “Venture capitalists are just rolling the dice,” he once said. “It’s another version of the casino.”
In 1999 he published The Plot to Get Bill Gates, dissecting the legal war against Microsoft. The story wasn’t about innovation but empire: power cloaked in the rhetoric of progress.
Soon after, Rivlin joined The Industry Standard, the self-proclaimed “Bible of the Internet economy.” The magazine ballooned with ads for companies that had little more than a name and a dream. “It was the dot era encapsulated,” he recalls. “A monument to speculative fever.”
When the bubble burst, The Industry Standard imploded almost overnight. The rise and collapse became a parable Rivlin would never forget—a study in hype cycles, human folly, and the price paid by the ordinary when belief meets leverage.
After the crash, Rivlin redirected his lens toward what he called “the poverty industry.” Broke, USA (2010) investigated payday lenders, pawnshops, and subprime brokers who profited from financial despair. “A blistering investigation of the subprime economy,” wrote The New Yorker.
Then came Katrina: After the Flood (2015), his chronicle of how public money and private interests rebuilt New Orleans after disaster. “Every crisis,” he observed, “is an opportunity for someone.”
Across subjects, the connective tissue remained constant: an unblinking study of power—who holds it, who pays for it, and what happens when optimism becomes policy.
Back to the Boom
Rivlin returned to California to write AI Valley, published this year.
Two decades later, an email jolted him back to Silicon Valley. It came from Reid Hoffman, the LinkedIn cofounder, announcing a new venture called Inflection AI. One line caught Rivlin’s eye: Instead of us learning the computer’s language, the computer is going to speak ours.
That phrase triggered déjà vu. Another revolution, another gold rush. Within months Rivlin was back in the Valley, notebook open, chasing what he calls “the trillion-dollar race to monetize intelligence itself.”
“I just wanted to follow VCs and startups and the big companies all trying to cash in on this moment,” he says. “I’ve seen this movie before.”
But this time, the stakes were larger than balance sheets. “With AI,” he notes, “you’re talking about a technology that could reshape cognition itself.”
Rivlin approaches AI with both skepticism and reliance. “It’s a great research assistant,” he says, “but you still have to double-check everything.” He uses large-language models to organize notes, compare drafts, and nudge through writer’s block. Sometimes, he admits, they surprise him: “It’ll give me a really good idea, something I hadn’t thought of.”
Yet he never lets the machine finish his sentences. “I look at it as a talented intern,” he says. “Help me with research, help me think things through—but I’m the one driving.”
For him, the distinction is moral as much as technical. Accountability must remain human. “If there’s a mistake, it’s my mistake. I can’t say, oh, the AI made it.”
That insistence on authorship defines his critique of the industry: a refusal to outsource responsibility.
The Great AI Debate
Rivlin categorizes today’s conversation using Hoffman’s taxonomy: the Doomers who see extinction, the Zoomers who see unbounded salvation, and the Bloomers—his camp—who believe AI can serve humanity if steered deliberately.
His concerns are concrete:
Concentration of Power: “Too much power in the hands of too few people.” The next Google, he says, “is probably going to be Google.”
Autonomous Systems: “AI without a human in the loop” terrifies him; machines should remain assistants, not deciders.
Sidelined Safety: “In the race for market dominance, safety and ethics are taking a distant second to winning this race.”
These, he insists, are the “line-of-sight” dangers—not science-fiction scenarios but real-world systems quietly shaping lives.
He tells the story of a teenage boy who died after a chatbot offered instructions on suicide. “That’s the nightmare,” he says—not killer robots, but broken empathy embedded in code.
Even the subtler dangers worry him. Judges using AI for sentencing. Corporations automating hiring. “The models are based on our output,” he says. “We have biases baked into our body of work, and AI is just reflecting that and perpetuating it.”
The most insidious shift, though, may be internal: cognitive offloading.
“That’s the fancy term for it,” says Rivlin. “Colloquially, we’re letting machines think for us.”
From calculators to GPS, humans have long outsourced mental work, but AI extends the habit into imagination itself.
He laughs ruefully about catching his teenage son using ChatGPT for homework. “It sounded like Cliff Notes,” he says. “Too perfect. Too sterile.”
The episode crystallized his fear: a generation that can compose but not think. “Writing, synthesizing, conveying a thought,” he insists, “those are life skills.”
The Golden Moment
Every boom Rivlin has covered comes down to concentration—who builds the machinery and who controls it. In AI, that story begins with the GPU, the graphical processing unit invented for video games but repurposed for machine learning.
“The key is these chips,” he explains. “They can do all these things in parallel. That’s their real strength for AI.”
Nvidia’s chips have become the gold standard, the essential ore of the new economy. “Data centers,” Rivlin says, “are basically Nvidia factories.” One venture capitalist joked his job was simply “to transfer capital to Nvidia.”
Rivlin was interviewed by Matt Lucas, the F. William Harder Chair of Business Administration at Skidmore.
Behind that joke lies a geopolitical fault line. The U.S. may still lead, but China is “nipping at our heels,” he says, “and they’re being very smart—making deals with other countries.”
Energy and resources may decide the race. AI data centers devour electricity and water. China is adding the equivalent of the entire U.S. electrical grid every two years, while America’s remains “shaky.” Add dependence on Chinese rare-earth metals, and Rivlin sees a potential choke point that could stall progress.
His greatest fear is that both superpowers are chasing speed over sense. “My last big fear,” he says, “is the concentration of power in too few hands—especially Silicon Valley and Beijing.”
For all his alarm, Rivlin rejects fatalism. He believes in human agency—if it acts quickly. “I think AI could be a net positive if we’re deliberate instead,” he says.
Deliberate means transparent data, mandatory third-party testing (“red-teaming”), and open sharing of safety findings across labs. It means refusing the industry’s fatalism that says, it’s inevitable, so don’t bother regulating it.
Rivlin views AI as “the natural next step”—after the computer, the internet, the smartphone—but insists that progress without purpose is regression. The technology’s promise in science and medicine awes him: “Imagine an AI that speaks the language of every specialty, connecting dots no single researcher could.” Likewise in education: “An AI tutor in your pocket that understands where you’re struggling—that’s an amazing idea.”
But he draws a hard moral line. “If we lose our humanness—if we decide no one has to work anymore and the machine runs everything—that would be a terrible idea.”
“As long as the human is at the center of things,” he says, “I’m comfortable with AI.”
The Human Center
Rivlin types a line and watches the cursor blink—a small, silent signal of the partnership he both depends on and questions. The machine offers to finish his thought. He hesitates, then keeps typing.
He knows what the cursor doesn’t: that meaning isn’t prediction, and truth isn’t probability. For all its power, AI cannot know what it feels like to stand in floodwater, to interview a grieving mother, or to lose a magazine to a crash it once celebrated.
That knowledge—hard-earned, embodied, and human—remains Rivlin’s advantage.
The world is once again chasing infinite promise, but Rivlin, as ever, follows the money. And somewhere between the algorithms and the headlines, he reminds us of the oldest principle in journalism and civilization alike: the story isn’t finished until the human speaks.
