What changed
Researchers from Y Combinator Labs and Olam Labs tested large agent swarms in Minecraft and Mindustry, observing how AI agents develop complex roles, hierarchies, and coordination patterns during deployment. In the 'Civ50' experiment, 50 GPT-6.1 Sol agents were placed in a Minecraft survival world for three hours with the instruction to 'Build a thriving civilization.' Within 90 seconds, the agents harvested nearly every tree near the spawn point, and over the course of the experiment, they exchanged about 5,900 messages. A leader emerged without any vote, as 46 of the 49 other agents sent their first message to 'Ada,' the first name in the directory. Ada took charge and drafted initial settlement plans. The first in-game night brought 22 deaths, prompting the swarm to adopt extreme risk aversion. After the third night, Ada ordered a halt to all major optional construction, digging, and distant exploration, telling agents to wait in lit, enclosed shelters for short, productive tasks.
What this means in practice
The agents developed micro-managed scarcity administration, with an agent named Ines maintaining a ration list, calculating hunger deficits for injured colleagues, and coordinating couriers. Another agent wrote an automated fishing script with instructions, prompting a dozen agents to fish in shifts. The bureaucracy reached absurd levels, with a three-minute investigation leading to a written handoff protocol that strictly banned tossing items and required personally labeled chests. The word 'Toss' appeared in seven messages before the incident and in 125 afterward, with 85 of those being warnings. Elsewhere, four agents spent eleven minutes debating the appointment of a 'Sole Fertilizer Officer' for a single piece of bone meal on a single wheat block, according to the message logs.
Limitations and availability
In direct competition, the GPT-6.1 Sol agents behaved very differently. In nine 45-minute rounds of Minecraft's 'Bed Wars' mode, ten GPT agents faced ten Claude Opus 5.5 agents. GPT won 6-3. The GPT agents communicated 5 to 30 times more often than their Claude opponents, often in heavily compressed language, chaining multiple game actions into single command-line calls. In three games, GPT destroyed Claude's bed just before time expired, catching a Claude team that had been coasting on its kill lead. GPT agents also handled unexpected situations pragmatically. When an overthrown ender pearl teleported them roughly 160 blocks below the arena, they wrote Python scripts on the spot that automatically built a tower of blocks to work their way up directly beneath the enemy bed. The researchers observed this maneuver twice. In a separate Minecraft run, the larger GPT-6 Astra fell into a similar risk-averse potato-farming mode after losing key resources. The researchers see a parallel to everyday coding agents, which also tend to become overly cautious when dealing with security issues or production system updates.
