Cursor demonstrated that cheaper AI models can handle most coding tasks when frontier models serve as planners. The company's upgraded agent swarm successfully rebuilt SQLite in Rust with 100 percent test accuracy, while the previous system failed due to merge conflicts.
Cursor tested its new agent swarm architecture against the old one in a challenging task: rebuilding SQLite in Rust using only documentation, with no source code or internet access.
The upgraded system separates roles between planner and worker agents. Frontier models handle planning and high-level decisions, while cheaper models execute the actual coding work. Every configuration of this hybrid approach achieved 100 percent accuracy on the test suite.
The predecessor system, which lacked this hierarchical structure, struggled significantly. It became trapped in merge conflicts created by its own changes, unable to resolve them without human intervention.
The results suggest a cost-effective path forward for AI-assisted development. By reserving expensive frontier models for complex reasoning and delegating execution to cheaper alternatives, developers could reduce inference costs while maintaining quality output. The architecture appears particularly effective for multi-step programming tasks that require both strategic planning and implementation.
The White House plans to expand its AI oversight framework in the coming months to include open-source models once they reach frontier capabilities, according to sources speaking with WIRED.
Anthropic's Claude chatbot is now accessible directly from a Chrome browser sidebar through the company's extension, eliminating the need to switch tabs for AI conversations.
Thousands of Indian workers are helping robotics companies collect training data by performing tasks like shoe stitching and steel welding. The videos will teach machines to automate these jobs.
RuntimeWire, an AI-powered news website operating on a $100 daily budget, scooped mainstream journalists—including WIRED—on an OpenAI security disclosure at Black Hat. The achievement signals a shift in how news breaks.