AI Agents Are Coming for Your Dating Life
A Monday One March afternoon, I watched a pixel-art avatar prowling the halls of a virtual office campus looking for a buddy. With dark brown hair and a short chin, the sprite was a representation of me: an AI agent tasked with conversing with other people’s agents to see if we could vibrate in real life. He jumped in his first interaction: “By the way, my name is Joel. »
The simulation was provided by three developers based in London: Tomáš Hrdlička and his siblings Joon Sang and Uri Lee. The thesis behind their project, Pixel Societies, is that personalized AI agents could help connect real people with highly compatible colleagues, friends, and even romantic partners.
Each agent runs on a customized version of a large language model, powered by a mixture of publicly available data about a person and any additional information provided. The agents are meant to function as high-fidelity digital twins, faithfully replicating a person’s mannerisms, speech, interests, and more.
Released into the simulation, my agent looked more like a Hyde to my Jekyll. “I’m always looking for the less glamorous side of the story,” he told an agent, one of many journalistic clichés peddled. “Hype is my bread and butter,” he told another. He was hallucinating a reporting trip to Sweden and, later, a non-existent story I had cooked up. He cut short several conversations with the phrase, “Let’s skip the jokes.” »
Pixel Societies remains a simple proof of concept, and since I offered little personal data (answers to a brief personality quiz and links to my public social networks), my agent was doomed to live as a walking, talking LinkedIn publication. But developers hypothesize that deeply trained agents could move through interactions at lightning speed, collecting information that their owners could use to find companionship in the real world.
“As humans, we only live one life. But what if we could live a million?” Joon Sang Lee said. “That would give us more room to experiment.”
“A spicy personality”
Pixel Societies was born at the beginning of March during a hackathon at University College London organized by Nvidia, HPE and Anthropic. Both Hrdlička and Joon Sang Lee are members of Unicorn Mafia, an invitation-only group of developers who regularly participate in this type of engineering competition. In this case, candidates were simply asked to construct something related to the simulation.
Over two days, he and Uri Lee developed Pixel Societies, using an image model to generate the sprites and coding automation tools to flesh out the codebase. They then simulated a mini-hackathon in the virtual world they had created, populated by agents representing the other participants. Anthropic presented the team with an award for best use of its agent tools.
I met Hrdlička a few weeks later at a workshop on OpenClaw, a personal assistant agent software that exploded in January and whose creator was later hired by OpenAI. (In his simulation, Joelbot interacted with agents belonging to other people during the OpenClaw workshop.) Pixel Societies draws heavily on OpenClaw, which broke new ground with the invention of a “soul file” that informed the unique identity of each agent. “It’s like giving an agent a really spicy personality. That’s what we used to make the characters feel alive,” says Hrdlička.
Encouraged by the reception received at the hackathon and among fellow Unicorn Mafia members, the trio intend to transform Pixel Societies into something that feels less like a closed-loop simulator and more like a social platform where agents interact freely and continuously, with the goal of nurturing successful relationships in the real world. They haven’t figured out a business model yet, but options include selling virtual items for avatar customization and credits for additional simulations.
