This week saw major progress in multi-agent AI systems where groups of specialized AI helpers work together. Amazon introduced AgentCore at its New York summit, calling it a "tectonic change" in software. This toolkit includes seven services helping businesses safely deploy AI helper teams at large scale. Amazon also pledged an extra $100 million to support development of these systems.

OpenAI launched a powerful new Agent mode for ChatGPT that can take real actions for users. It can shop online, make reservations, and plan events like ordering wedding outfits based on weather and dress codes. This feature rolled out on July 17 but isn't available in the European Union due to regulatory concerns, frustrating users there.

In healthcare, Codex Genetics presented a multi-agent system at a medical conference in Asia. Their AI team helps match cancer patients to suitable clinical trials faster than before. Customer service centers are adopting similar systems where different AI helpers specialize in tasks like security checks, troubleshooting, and sales - working together like an ant colony for smoother service.

Workforce experts explain that these AI teams will change jobs, not replace them. Humans will focus on setting goals, making ethical decisions, and handling creative tasks while AI manages routine coordination. Important questions about trust, control, and accountability remain when AI makes independent decisions.

Investment in this technology is growing rapidly. Industry reports highlight 15 leading companies developing agentic AI, including Microsoft and Amazon who are creating specialized teams within their companies. Researchers at the University of Birmingham hosted workshops exploring how groups of AI helpers could assist in public decision-making by representing different viewpoints.

As these systems spread, companies are creating marketplaces for pre-built AI helpers and publishing guides to help developers build their own agent teams. The race is on to develop secure frameworks that let businesses harness the power of multiple AI helpers working together while addressing safety concerns.