Michigan prof creates Python-like Jac language for AI agents
Jason Mars unveiled Jac and Jaseci at the All Things Open conference, arguing that existing programming languages were designed for humans and poorly suit AI agent development. Mars cited the example of Jarred Sumner spending $165,000 in API costs to convert Bun framework from Zig to Rust using an AI model, arguing this expense reflects the mismatch between human-centered tools and agent needs. Jac represents a shift toward tooling optimized for AI agent workflows.
Key points
- Jac language treats AI agents as first-class programming entities
- Designed to reduce complexity and cost of agent-driven software development
- Addresses inefficiencies of using human-centered tools for AI workflows
Summary and key points written by AI from the headline and a short excerpt, so they can be incomplete or wrong. The Register did the reporting; read the original for the full story.
Read the full story at The Register
