TheAgentCompany (CMU Research)
https://github.com/TheAgentCompany/TheAgentCompanyCarnegie Mellon University's benchmark simulation of a company staffed entirely by AI agents — revealing that the best models complete only 24% of real business tasks.
Carnegie Mellon University's benchmark simulation of a company staffed entirely by AI agents — revealing that the best models complete only 24% of real business tasks.
Research initiative and experimental platform exploring fully autonomous organizations where AI agents perform all work including decision-making and execution.
Building the 'Safe Autonomous Organization' — ran a real-world experiment with Anthropic where an AI agent autonomously operated a vending business in San Francisco.