I entered the library of a Park Avenue mansion, surrounded by books on polar exploration and model ships. Alone among a crowd of innovators and entrepreneurs, I felt anxious and out of place.

Why was I uptown at the Explorers Club for an artificial intelligence conference instead of downtown at work, teaching Spanish to teenagers?

In truth, I went to support a close friend who had organized the event and learn more about his work. I also wanted to feed my own curiosity about AI and wrestle with my own ambivalence.

Could AI help me learn and grow as a writer, teacher, and human? Or was it an existential threat? Could AI help me pursue a life of purpose and service? Or was it a digital distraction, an opiate for the masses meant to enrich billionaires and corporations? Could AI actually make me more intelligent? Or would it render me a mindless zombie addicted to screens, a Gen X version of The Anxious Generation?

The club exuded a low-key exclusive vibe, familiar from my time in prep school, college in New Haven, and countless British period dramas. Alone and awkward, I paced a wood-paneled library. Shelves crammed with tomes on Shackleton’s Antarctic adventures reminded me of the club’s elite ranks, which include Neil Armstrong, Sir Edmund Hlilary, and Jane Goodall (RIP), as well as Jeff Bezos and James Cameron.

Entering the main conference room, I spotted a familiar face: my dad. Since he works in tech, I’d suggested he come. We chatted for a few minutes, then I excused myself to another table. We could talk any time. Today, we were here to learn from new people.

Like my friend Jeff Wilser, an author and journalist who hosts the podcast AI Curious, and conference co-creator Jennifer Strong, I was curious about the promise and perils of artificial intelligence. Would AI enable healthier and happier humans and save the planet from destruction? Would it spark a sci-fi apocalypse, with humans as slaves to robot overlords? Or was there a sweet spot between AI hype and hysteria?

As I settled in, I spotted another friendly face: Dolly Chugh, a superstar professor whose two books I had helped edit. We hadn’t seen each other since she visited the creative writing class I was teaching at a state prison. We smiled and waved, and then the conference kicked into gear, an immersive day of learning and connecting.

I had feared a parade of tech bros preaching digital salvation. Instead, I found a philosopher, a magician, and a surprisingly diverse crowd of speakers and attendees who spanned sectors, generations, and vibes.

The Perils of AI

As opening speaker Reid Blackman warned, AI is an imminent disaster: too sprawling and complex for individuals or institutions to control and evolving at an alarming rate. To mitigate risk, the ex-philosophy professor turned corporate consigliere urged business leaders to quit dreaming, imagine their worst AI nightmare, then act accordingly. While he did not specify what that might entail, his candor was refreshing, his bluntness comforting. Naming a problem is the first step to taming it. His framing made me wonder what problems I had yet to name and tame.

Other stories of combating AI “slop” and “hallucinations” reminded me of the technology’s peril. While large language models have phenomenal power to synthesize information, they are also prone to making phenomenal errors. As a popular AI app, warns in the fine print at the bottom of the screen:

Chat GPT can make mistakes. Check important info.

That neutral disclaimer understates the risk. AI watchdog Hilke Schellmann, author of The Algorithm: How AI decides who gets hired, monitored, promoted and fired and why we need to fight back right now, shared shocking stories about how AI can harm employers and job seekers alike by amplifying biases and making egregious errors.

Beyond employment, her work reminded me: AI errors can be life and death. For example, AI has profound implications for medicine and the military. On the one hand, AI can help doctors diagnose patients and generals wage wars. But when the machines make mistakes, people die.

When Francesco Rulli, a dapper Italian entrepreneur shared tales of his digital twin, a virtual clone that deals with his drudgery while he practices judo and sails the Mediterranean, the tool sounded tantalizing. Later, the invention also suggested an episode of Black Mirror or Severance: a dystopian future in which humans have effectively enslaved machines.

The Promise of AI

While I didn’t follow all the scientific jargon, I was inspired by panels on how artificial intelligence was accelerating advances in biology and chemistry and facilitating discovery and production of medicines, a notoriously expensive and time-consuming process, with massive implications for human health.

To this non-scientist, it seems plausible that AI could accelerate drug research and discovery. But since the slowest part of drug development is clinical trials, the larger question seems to be: Could AI replace human trials? And if so, would we be willing as a society to take medicine that hadn’t been tested on humans?

Similarly, I was intrigued by how Altana was using AI to streamline global trade, connecting buyers, suppliers, and governments. This logistical application of AI, along with panels from Citibank and Neiman Marcus, reminded me that not all AI uses are glamorous; beneath the hype, AI is fundamentally a tool, if a powerful one. When Altana’s founder likened the advent of AI to the arrival of electrification, the analogy seemed apt, not hyperbolic.

Later, I was moved and amused by several young founders who pitched their start-ups.

Two babyfaced engineers invented AI robot hands to help astronauts load cargo, and had contracted with the international space station. Their joystick-powered product reminded me of arcade games where you try to grab a stuffed animal with a mechanical claw.

Later, I registered their company name, Icarus Robotics, an unsubtle allusion to the Greek myth about a fabulous invention that fails. Were they acknowledging the risk of flying too close to the sun–or mocking the moral about the limits of technology and the cost of human arrogance?

Another young founder was more down to earth. Clad in a black hoodie with white letters that said F*CK ACL TEARS, he had wanted to be the next Jeremy Lin until a knee injury destroyed his college basketball career. Since then, he had invented an AI-powered knee brace that inflated automatically like a car’s airbag. and was moving to New York to launch Hippos Exoskeleton, his fledgling company.

With their earnest energy and precocious enthusiasm, the young founders seemed straight out of Silicon Valley central casting. They also reminded me of many students I’ve taught and gave me a glimpse of a potential future for my son. Their presence at the conference also served as a tacit reminder to the mostly middle-aged crowd: innovation, especially technological innovation, is a young person’s game.

Cultivating connection

If I’d wanted only to hear people talk about AI, I could have stayed home and listened to AI Curious, my friend Jeff’s podcast. For me, a surprising joy of the conference came from small moments of human connection.

In the morning, I chatted with an ex-architect whose company used AI to make special effects for movies, and the author of a book on women in the workplace.

Over lunch, I mingled with an intellectual property attorney and an executive coach. We bonded over the craziness of AI-generated music and briefly used an AI app to compose a theme song for the conference, which was fun until the app got glitchy and threw up a paywall.

In the afternoon, I bonded with a publicist, an organizational psychology professor, and a brand strategist who moonlights as a comedian who pits humans against AI in live joke battles.

At the cocktail reception, I re-connected with three people I had actually known before the conference: my friend, my dad, and Dolly, now writing a new book about humane leadership. I also thanked the magician whose lighthearted closing act had combined old-fashioned mind-reading and card tricks with an AI twist. Later, I appreciated the curation: ending a day of dense discussions with magic and laughter.

When Deepak Agarwal, LinkedIn’s chief AI officer, said he viewed AI as a tool to increase connection between humans, it sounded like he was promoting his employer. a job search platform with 1.2 billion members, that my table companion wryly dubbed “social media for adults.”

Still, the conference exemplified his point: AI had inspired a group of people who might not have otherwise met to connect in person. If I had one regret about the day, it was that I didn’t meet more people or make deeper connections.

Before the salon, I thought I wasn’t an AI person. After, I realized we’re all AI people, whether we like it or not.

Back to school

As a teacher, I enjoyed spending an entire day at the AI salon in the role of student. And as a non-business person, I enjoyed learning about the impact of AI on business. Still, I do wish someone had addressed the profound impact of AI on education. On the one hand, AI is a powerful tool for both teachers and students. On the other hand, it poses an existential threat to the enterprise of school. And obviously the way we educate students has profound implications for the future of business and society.

When I was teaching writing at NYU, it was clear that college students were using AI to write papers. Meanwhile, professors were struggling to adjust to the new reality, cycling through their version of the seven stages of grief. Likewise, my colleagues and friends who teach college and high school students have been wrestling with their students’ use of AI. And while AI impacts all academic disciplines, its most palpable impact is in the humanities, where students are required to write.

Beyond questions of cheating in school, AI invites a deeper reflection about the purpose of education? What do we want students to learn and why? And since AI is part of the problem in education, could it also be part of the solution?

Back to work

As I left the Explorers Club that evening, I felt energized and inspired. Would I have enjoyed the event as much if it hadn’t been organized by a friend? Maybe. But like all humans —and all AIs—I’m biased.

After the conference, I had more questions than answers. But maybe that was the point of the immersive salon. As IBM chief scientist Dr. Ruchir Puri opined: “Knowing what you know is retrieval. Knowing what you don’t know is intelligence.” The goals he outlined for AI—consistency, awareness of limitations, and the capacity for self-improvement–echoed my own aspirations as a writer, teacher, and human.

Back at work in the morning, I would go back to teaching students to speak Spanish, not master neural networks. We would not marvel over scientists and entrepreneurs, but icons like labor activist César Chávez, who marched 300 miles and waged peaceful protests and hunger strikes to fight for farmworkers’ rights, and Lorena Ramírez who ran 100K ultra-marathons in plastic sandals to support her indigenous community in rural Mexico.

Still, the essential questions–in the classroom and at the conference–were the same.

What do we need to learn to thrive and survive in a volatile world? How can we leverage language to understand ourselves and each other? How can we cultivate curiosity and connection?

When I posed these three questions to Chat GPT, the generative AI response was:

We need to listen deeply, speak thoughtfully, and stay endlessly curious—building understanding and connection wherever we go.

I totally agree. But to my ear the language sounds canned and lacking in humanity. Platitude City.

And when I entered the opening paragraph of this essay into an AI image generator, the illustration of a robot in a library felt obvious and cold, lacking the warmth of the spontaneous phone photo I snapped of Magic Mike trying to read Dolly’s mind.

Maybe one day the machines will make all our words and images so we can go sailing or swimming or whatever.

For now, we still have work to do.