Artificial intelligence agents are coming to your dating life

Monday On a March afternoon, I watched a pixelated avatar roam the hallways of a virtual office campus looking for a friend. With dark brown hair and a jagged chin, the goblin represented me I have an agent We were instructed to talk to other people’s agents to see if we could feel the vibe in real life. He jumped into his first interaction: “By the way, I’m Joel.”

The simulation was run by three London-based developers: Tomáš Hrdlička and siblings Joon Sang and Uri Lee. The thesis behind their project, Pixel Societies, is that of personality Artificial intelligence agents It can help match real-life people, colleagues, friends, and even highly compatible romantic partners.

Each agent runs on a customized version of a large language model, fed with a combination of publicly available data about the person and any additional information they provide. Agents are supposed to act as highly accurate digital twins, faithfully imitating a person’s style, speech, interests, etc.

If we’re allowing simulations, my agent was my Hyde to my Jekyll. “I’m always looking for the less glamorous side of the story,” she told one client, one of her many journalistic clichés. Another said: “Noise is my daily bread.” It was a hallucination during a press trip to Sweden, and afterwards, a non-existent story that I was said to be cooking surfaced. He has summed up several conversations with the phrase, “Let’s skip the pleasantries.”

Pixel Societies remains a basic proof of concept, and because I provided little personal data — responses to a short personality quiz and links to my public-facing social media — my agent was doomed to life as a walking, talking LinkedIn post. But the developers assume that deeply trained agents can navigate through interactions at breakneck speed, gathering information their owners can use to find companionship in the real world.

“As humans, we only live one life. But what if we could live a million lives?” Joon Sang Lee says. “It will give us more breadth to experiment.”

“Spicy personality”

Pixel Societies was founded in early March at a hackathon at University College London hosted by Nvidia, HPE and Anthropic. Hrdlička and Joon Sang Lee are both members of the Unicorn Mafia, an invitation-only group of developers who regularly compete in this type of engineering competition. In this case, the contestants were simply asked to build something related to the simulation.

Over the course of two days, in collaboration with Ori Lee, they developed Pixel Societies, using an image model to create sprites and coding automation tools to flesh out the code base. They then simulated a mini-hackathon within the virtual world they created, featuring agents representing the other contestants. Anthropic gave the team an award for best use of its agent tools.

I met Hrdlička two weeks later at a workshop on OpenClaw-Agent personal assistant software It exploded in January whose creator was later assigned by OpenAI. (In his simulation, Joelbot interacted with agents belonging to other people in the OpenClaw workshop.) Pixel Societies draws great inspiration from OpenClaw, which got its start by inventing a “soul file” that identified the unique identity of each client. “It’s like giving the client a really interesting personality. That’s what we used to make the characters feel alive,” Hrdlicka says.

Encouraged by the reception at the hackathon and among fellow Unicorn Mafia members, the trio intends to turn Pixel Societies into something less like a closed-loop simulation and more like a social platform where customers interact freely and continuously, with the goal of sparking fruitful relationships in the real world. They haven’t come up with a business model yet, but options include selling virtual items for avatar customization and credits for additional simulations.