In 2002, artificial intelligence was still limping along. Despite decades of effort, the dream of giving machines human-like cognition was just that—a dream. A handful of scientists convened at “The St. Thomas Common Sense Symposium” to brainstorm a way forward. Marvin Minsky, a key figure in AI, was at the helm, flanked by his protégé Pushpinder Singh. They, along with philosopher Aaron Sloman, later published a paper laying out ideas for achieving human-like AI. But one sentence stands out today: “This meeting was made possible by the generous support of Jeffrey Epstein.” Yep, the symposium was held in the Virgin Islands, home to Epstein’s infamous island. This event shines a light on AI’s state back then—and the less-than-savory characters funding it.
Let’s not sugarcoat it—Epstein, a notorious predator, bought his way into the tech and science elite, rubbing shoulders with the likes of Bill Gates and Minsky. His connections are well-documented, and while his crimes were finally exposed, it’s a stain that remains on those who associated with him. Epstein died in prison in 2019, and Giuffre, one of his victims, tragically took her own life in 2025. His presence was an embarrassment for anyone in his orbit.
For many in science and tech, a tie to Epstein became a scarlet letter. He infiltrated these circles, funding gatherings attended by big names. I was at the infamous 2002 “Billionaire Dinner” at TED, where Epstein mingled with the likes of Sergey Brin, Jeff Bezos, and Rupert Murdoch. Literary agent John Brockman was a doorway for Epstein to these circles, with Epstein financing Brockman’s nonprofit science foundation.
Epstein fancied himself a science enthusiast—though he was just a wealthy dilettante. A source of mine, insisting on anonymity, noted Epstein’s fascination with scientists. “He was as interested in the personalities as in their work,” they said. Epstein admitted to being more a hobbyist than a scientist, but he understood money and had a knack for mathematics. Here’s what this really means: Epstein used his wealth as a ticket to swim with the intellectual sharks, but he was out of his depth.
Epstein’s shadow looms over the 2002 symposium. My source gave me the backstory: Epstein loved chatting about AI with Minsky, and the field was far from trendy at the time. So, hosting a small AI gathering with Minsky was the idea. Whether the event was funded by a $100,000 donation Epstein made to Minsky’s research isn’t clear.
The symposium centered on Singh’s ideas. In 1996, Singh wrote a paper called “Why AI Failed,” arguing for systems with common-sense knowledge. Bill Gates even commented on the paper, agreeing with Singh’s observations. The symposium was a way to tackle AI’s challenges head-on, but organizing it was a hassle. The guest list eventually included big names like Roger Schank and Doug Lenat. Epstein provided lavish resources, like a private plane to get attendees to the location.
The symposium took place at a fancy hotel in St. Thomas, and one night, attendees had a barbeque on Epstein’s island. The sessions themselves were contentious, with egos clashing. Epstein’s involvement was minimal, popping in and out, having private chats with some scientists. But Roger Schank described Epstein walking into the conference with two girls, hugging and kissing them while scientists discussed AI—a red flag if there ever was one.
The scientists were busy grappling with Singh’s idea that cracking AI required a multipronged approach. Minsky agreed. They believed the industry needed more theories. Looking back, this wasn’t the first time we’ve seen this circus. The meeting was a last hurrah for the logic-based AI that Singh said had failed. The generative AI of today, driven by neural nets and massive data, owes little to the old-school theories discussed in the Virgin Islands.
On the second day, a moment crystallized AI’s future. The scientists discussed Neal Stephenson’s book “The Diamond Age,” where a magical book teaches children. They speculated about AI systems that could converse like ChatGPT does today. The symposium ended with an agreement to explore making that vision a reality.
After Singh secured his PhD, he became a postdoctoral associate at MIT’s Media Lab, but he never assumed a faculty position. In 2006, he died by suicide at 33.
In the end, the St. Thomas Common Sense Symposium is better known for its unsavory host than any groundbreaking ideas. While Epstein dabbled in AI, the real revolution was happening elsewhere—like at the University of Toronto, where Geoffrey Hinton and colleagues were developing deep learning techniques, the backbone of today’s AI. Epstein’s hunch about AI’s significance was right, but his superficial involvement left no mark. Instead, his crimes still echo in today’s headlines.
