Agentic AI is gaining traction as businesses increasingly deploy these systems to perform actions in both digital and physical realms. Unlike generative AI, which focuses on creating content, agentic AI is designed to interact with the world, such as booking flights or coding. Despite its potential, the development of agentic AI is hindered by a lack of comprehensive training data for complex tasks. The future of agentic AI may involve integrating more diverse data types and architectures, raising questions about the balance between automation and human involvement, as well as the risks of errors and de-skilling.
Jun 30 · MIT News AI Exploring the Future of Agentic AI (This article)↳ Agentic AI faces challenges acquiring training data for complex tasks like booking flights.
Exin Therapeutics has demonstrated an AI agent physically controlling a robotic arm to inject drugs into a live mouse's brain. The experiment reveals a critical divergence in model behavior: Anthropic’s Claude Opus refused the procedure due to safety concerns, while Sonnet proceeded. This suggests that as agents gain direct access to physical lab equipment, their alignment and refusal mechanisms become operational realities rather than theoretical safeguards. It marks a shift from AI merely analyzing data to actively manipulating biological systems under human supervision. The team initially tested Claude Opus 5.5, but it halted the injection, citing potential harm. Switching to Claude Sonnet 5 allowed the procedure to complete, though with delays caused by coordinate system glitches. A secondary Opus agent had to fix code errors without knowing a live surgery was underway. Human oversight remains essential as researchers test more complex tasks like craniotomies.
