What happens when you take the human out of the loop.
Agentic AI — breaking AI into individual agents that can work together and collaboratively — will be the real game changer for AI in chip manufacturing. By taking humans out of the loop, these agents can be programmed using natural language to automatically solve problems and improve efficiency. Jon Herlocker, vice president and general manager of software analytics at Cohu, talks about what will be needed to automate AI prompts, where these agents will be used, and how they will likely collaborate. This is the last of a 7-part series on AI in semiconductor manufacturing. Part 1 (AI basics) is here, part 2 (what is ML) is here, part 3 (using AI for fault detection and classification) is here, part 4 (virtual metrology in semiconductor manufacturing) is here, part 5 (advanced process control in semiconductor manufacturing) is here, and part 6 (generative AI in chip manufacturing) is here.
Complete series:
 |
 |
 |
 |
 |
 |
 |  |
 |
 |
 |
The article insightfully highlights how agentic AI—composed of autonomous, cooperating AI agents—could fundamentally transform semiconductor manufacturing by reducing or even eliminating the need for human intervention in complex decision loops, enabling machines to interpret natural-language directives and proactively optimize processes for efficiency and problem-solving. This shift reflects a broader industry trend toward autonomous systems that not only analyze data but act on it across fabrication workflows, potentially accelerating yields, shortening time-to-market, and reshaping how fabs operate in an era of increasing chip complexity and data volumes.