OpenClaw founder Peter Steinberger operates 100 AI agents powered by OpenAI's Codex, generating $1.3 million in monthly API costs. The three-person team frames the expenditure as research into software development when token constraints are removed.
Steinberger's experiment explores a fundamental question: how does the development process change when AI coding costs become negligible?
The 100 Codex instances handle core development tasks including writing code, reviewing pull requests, and identifying bugs. By operating at scale without budget constraints, the project generates data on AI agent efficiency and coordination in real-world software development scenarios.
The substantial monthly bill reflects OpenAI's current pricing structure for API access. For context, this spending level demonstrates both the computational intensity of running multiple large language model instances and the economics of AI-driven development at scale.
The OpenClaw initiative represents an emerging class of experiments that prioritize research insights over cost optimization. Rather than minimizing API spend, Steinberger's team maximizes agent deployment to understand optimal configurations for autonomous development workflows.
Results from this research could inform how organizations approach AI integration in software engineering as model costs continue to evolve.
Startups like Altur are deploying AI chatbots to handle debt collection calls, automating a process traditionally done by humans. Y Combinator has backed six debt collection and settlement startups over the past six years.
Following recent earthquakes, Venezuelan developers and citizens deployed AI-powered websites and apps to locate missing persons and coordinate disaster relief as government response lagged.
Prime Minister Anthony Albanese has created a dedicated AI office and committed to protecting Australian creators from copyright infringement by artificial intelligence companies. The government rejected plans to grant tech firms free access to Australian data.