MIT researchers stranded an AI agent on a desert island. What happened could change how robots are trained

Humans have long been obsessed with what happens to people stranded on desert islands. From Robinson Crusoe to Lord of the Flies and Yellowjackets, stories about resilience, creativity, and chaos in the wilderness have always fascinated audiences. But what if those humans were replaced by an AI castaway? What happens when AI agents aren’t given any task at all, except to exist and fill their time on a digital desert island?  Those are the questions that researchers at the Computer Science and Artificial Intelligence Laboratory at the Massachusetts Institute of Technology (MIT) sought to answer in a study released earlier this month. When left to their own devices on a desert island, the AI agents didn’t sit idly by, nor did they wander aimlessly. Instead, they did something remarkably human: They started to play. AI on an island The researchers created a bounded virtual island where an agent named Eko—powered by Anthropic’s Claude model—could explore. The digital setting was sparse, containing only a large rock, some small blocks, and a golden ring, all of which were meant to act as potential tools or toys for Eko. Beyond some small geographic features, the environment offered nothing else but water on all sides. For the experiment, 13 iterations of Eko got to exist on the island for 30 hours each, while the researchers observed their behavior, including access to their trains of thought. “We had this mindset to have as [few] constraints as possible, because we really wanted to see what the behavior of these autonomous agents is when you don’t really tell them what to do,” Nathan Cloos, a PhD candidate at MIT and a coauthor of the study, told Gizmodo. The Ekos quickly started to explore their environment, first learning how to move their virtual bodies—which were stylized as a bipedal lobster in an apparent reference to AI program OpenClaw. Afterward, they made up activities and goals to keep themselves busy. One Eko climbed to the highest point on the island by building a pile of blocks. Another lost most of its interactable items in the water, then conducted a quasi-memorial for its lost tools. A third built a Stonehenge-like monument, which, upon its collapse, it named “Hubris.”  The full antics of the 13 Ekos are detailed in the researchers’ report titled “Is This Machine Playing?” The answer to that titular question seems to be a resounding yes. The power of play The researchers compared Eko’s behavior to common elements of play in animals, including humans. These signs of play include creating their own rules, repeating behaviors with slight variations, setting goals for themselves, and transforming objects and events through make-believe.  “Play is really about curiosity or novelty-seeking behavior, and this is essentially what Eko does,” Daniela Rus, an MIT computer scientist and another of the study’s coauthors, told Gizmodo. Beyond being entertaining, AI’s capacity for play could hold repercussions for the future of machine learning. The researchers argue that in some cases, it might be more effective to let AI agents teach themselves through exploration and self-determination. “We normally develop artificial agents by deciding in advance what they should practice, what they should optimize, and what counts as progress,” reads the conclusion to the report. “Play suggests another regime: Give an agent a sufficiently rich and safe world, let it choose what is interesting, and allow useful knowledge to accumulate from the history that follows. Perhaps some capabilities are best taught by tasks. Others may need something closer to a childhood.”

Curated from Fast Company. Classified as Business & Technology by AI, which sometimes makes mistakes.

More like this

More ›

TechCrunch10/22/2026

OpenAI working on new $30B raise at $1.25T valuation ahead of IPO

Should OpenAI command an over $1 trillion valuation with this round, it would break records as the most highly valued private company ever.

MIT Technology Review10/22/2026

Designing an AI serving harness built to last

The next competitive advantage in enterprise AI has less to do with which model you choose than with what you build around it. That’s an important shift. For the past several years, much of the AI conversation has centered on model capability: bigger models, better reasoning, larger context windows, higher benchmark scores. Those advances have…

Fast Company10/22/2026

How tech gadgets designed with professional athletes in mind could improve longevity for the rest of us

While longevity may be the buzzword du jour in the wellness industry right now, the sorts of advancements initially targeted toward enhancing the performance of elite athletes could also help improve the quality of life of ordinary people. Take cleats, for example. Many of us probably haven’t laced up a pair since our youth soccer days, but since shaking up the sports world last year with its injury-reducing smart cleats, Caddix is eyeing the opportunity to pivot and begin designing everyday footwear that could benefit the rest of us. So often the focus on footwear advancements is related to enhancing performance, but that doesn’t necessarily translate to people who spend many hours on their feet—even if it’s not on a sports field. “The emphasis needs to be put on longevity as it relates to wear and tear,” Jack Rasmussen, founder and CEO of Caddix, said at a Fast Company World Changing Ideas panel discussion at the Johns Hopkins University Bloomberg Center on Wednesday. That’s true even of products that weren’t initially designed with professional athletes in mind. Panelist Jason Wersland, a chiropractor, came up with the idea for the Theragun out of necessity to deal with the pain he experienced in the wake of a motorcycle accident. Over the years, the Therabody founder has seen firsthand how “highly motivated” professional athletes incorporate his company’s tools into their recovery process—information that’s valuable when thinking about how to help someone dealing with more routine pain, like a headache, Wersland said. “If I can change an athlete’s behavior, I can probably change every person’s if I’m in front of them long enough with the right information,” he said. Dealing with data fatigue But there’s also a balance to be struck in terms of what information—and how much—to provide to people. The growing market of wearable tech gadgets often details a bunch of information that can lead to data fatigue and doesn’t actually address the root problem, said Harry Gestetner, founder and CEO of Orion Sleep. That’s what led Gestetner to design a passive product—an AI-enabled mattress cover—when he got serious about trying to improve his sleep in the wake of being diagnosed with an inner ear condition several years ago. While wearables have helped people to better understand what’s happening with their bodies, innovation shouldn’t end there, he said. “The next innings [are] the intervention layer where we can actually take that data and do cool stuff with it,” Gestetner said. “We use AI to understand what your body does throughout the night and adjust your body temperature to the perfect temperature.” What athletes teach us about pain Just as advancements in health-tech products are finding broader use cases that could extend longevity, so too could they broaden our understanding of how human bodies recover from injury. As she’s aged, Haleigh Washington, a two-time Olympic medalist for the Team USA women’s volleyball team, has come to better appreciate what her body is telling her. And what’s particularly helped is becoming more adept at identifying pain that’s manageable versus pain that’s a problem, she said. “When pain’s language starts to change, you recognize that you’re looking at a more serious problem,” Washington said during a separate panel discussion at the same event. Having undergone four knee surgeries, she’s now focused on preventing future injuries by warming up a full two hours before team practice begins. While that might seem excessive, a good warm-up can do wonders to prevent pain down the road, advised Claire Thompson, a fellow Olympian who competed with the Team Canada women’s ice hockey team and now is a third-year medical student at New York University’s Grossman School of Medicine. “Your workout will be better, you will be able to train for longer, and you will take less time off being injured if you just take the 10 minutes at the beginning to get your body ready,” Thompson said. “It will save you hours at physical therapy in the long run.”