Did Google’s AI agents really build an operating system for $916?

The importance of independent evaluation

By Stephan Rabanser, Sayash Kapoor, Rishi Bommasani, Andrew Schwartz, Arvind Narayanan At Google’s developer conference earlier this week, the company launched its latest model, Gemini 3.5 Flash, alongside a new agent app, Antigravity 2.0. To showcase what this new agent setup is capable of, Google claimed that a team of agents had built an entire operating system. The effort reportedly required only a single prompt, cost only about $900 in API fees, and was carried out by a few dozen subagents working together. Does this mean that complex pieces of software can now be built cheaply by AI?…

related reading

• Google "We Have No Moat, And Neither Does OpenAI"semianalysis.com
• Google Antigravity Blog: Google Antigravity Built an OS (and more)antigravity.google
• Demystifying evals for AI agents \ Anthropicanthropic.com
• Towards self-driving codebases · Cursorcursor.com
• Building Effective AI Agents \ Anthropicanthropic.com
• Building Effective AI Agents \ Anthropicanthropic.com
• Agent swarms and the new model economics · Cursorcursor.com
• Scaling long-running autonomous coding · Cursorcursor.com
• A basic systems architecture for AI agents that do autonomous research — LessWronglesswrong.com
• Why I'm Betting Against AI Agents in 2025 (Despite Building Them)utkarshkanwat.com
• Gemini 3 is Evaluation-Paranoid and Contaminated — LessWronglesswrong.com
• OpenAI researcher on agent swarms & recursive self-improvementyoutube.com