🧠 Food for thought

Implications, context, and why it matters.

Michelo’s strategy centers on keeping skilled workers’ judgment in the loop

• Michelo Robotics started with a simple question. Why do factory finishing jobs like painting along curves or sanding with steady pressure still depend on people 1?
• Its answer is that these tasks rely on “human sense,” not knowledge that can be fully written into code. That idea sits behind its mission, “Bringing Human Sense to Robot Intelligence” 1.
• The company says robots can learn and hold onto that “human touch” while handling repetitive or risky work. Its goal is “Human-Robot Harmony,” where machines grow alongside people rather than push them out 1.
• That approach puts worker know-how at the center and treats the company as a partner that learns from skilled craft, instead of focusing only on faster output 1.

Michelo focuses on physical skill while much of robotics chases language-based reasoning

• Michelo’s effort to capture manual expertise sits in a narrower AI lane. Across robotics, many teams are adding language-based reasoning 2, 3.
• GoodAI’s Prometheus drone swarms, groups of drones built to work together, use large language models. They are meant to act on broad commands such as “Search the second floor,” with the focus on autonomous reasoning rather than copying human movement 2.
• The IROS 2025 program, a major robotics research conference, also includes work on a multimodal large language model as a higher-level reasoning unit for robot-assisted surgery 3.
• As a result, Michelo is working to capture tacit physical knowledge, the know-how skilled workers use but may struggle to explain step by step, while much of the field is trying to build robots that can carry out complex language-based instructions 2, 3.