Great Question: A Manufacturing Podcast The AI You Don’t Want To Use (IndustryWeek)
Feb 26, 2026
Pradip (Pradeep) Singh, Chief Manufacturing Officer at GlobalFoundries with decades in semiconductor operations, explains why most AI pitches miss the mark. He describes being inundated with unrealistic proposals. He stresses AI should augment operators, not replace them. He highlights the need for manufacturing domain expertise, warns against automating human management, and points to scalable, end-to-end AI as worthwhile.
AI Snips
Chapters
Transcript
Episode notes
Startup Claimed They Could Eliminate Fab Operators
- Pradeep Singh recalled a three-person startup claiming they would eliminate operators and cut manufacturing costs by 75 percent.
- He laughed, kept them long enough to see they couldn't deliver, and they never returned after overstating feasibility.
Learn The Factory Problem Before Building AI
- Do learn the customer's real problems before building solutions and avoid a 'hammer looking for a nail' approach.
- Pradeep coaches startups to identify manufacturing challenges and tailor pitches to what fabs actually need.
Most AI Startups Lack Real Manufacturing Experience
- Pradeep estimates fewer than 15 percent of AI pitches show real manufacturing understanding.
- Most teams are strong programmers but lack hands-on experience in large-scale fabs and miss domain constraints.