In the past few years manufacturers have asked a familiar question – “How can we use artificial intelligence?” But that question is quickly becoming too simple. A better, more strategic question is: “How can we make AI understand operations?”
This distinction is vital. Generic AI can summarize documents, draft emails, analyze spreadsheets, and generate ideas effortlessly. These are useful, time-saving outcomes but they rarely transformative on the plant floor. True manufacturing advantage will stem from something much more specific: AI that is customized to a manufacturer’s products, processes, equipment, workforce, suppliers, customers, and particular operating constraints.
The next phase of AI in manufacturing will not be merely about access but more specifically about “fit,” meaning how well AI understands the specific factory, equipment, people, and decisions it is meant to support. Fit measures how well AI understands the specific plant, machines, people, and decisions it is meant to support.
Generic AI falls short
A manufacturing plant is not a spreadsheet. It involves multiple different machines with quirky behaviors, operators with varying experience levels, complex maintenance histories, supplier delays, strict quality thresholds, safety regulations, and crucial customer commitments that generic models cannot fully grasp. A tool that helps one plant reduce scrap might be completely irrelevant - or even risky - in another.
So customization is needed, but it cannot lead to AI improvise freely. AI customization must happen strictly within boundaries defined by safety, quality, compliance, cybersecurity, and accountability.
The historical parallel
There is a compelling historical parallel in the electrification of manufacturing. Electric power did not instantly transform factories once it became available. Many plants simply plugged electric motors into operating models originally designed for steam power. Massive productivity gains did not materialize until manufacturers redesigned their systems around the new capability, moving toward flexible layouts organized around materials, people, and processes rather than centralized power transmission.
AI is entering a similar stage now. Many organizations are simply bolting AI onto legacy workflows to speed up reports or summarize data. These are reasonable starting points, but they are not the endgame. Real advantages will accrue to the companies that use AI's capacity for faster learning, adaptive decisions, and precise coordination to completely redesign their workflows.
First and second-order effects
AI is moving beyond screens into the physical realm of factories, warehouses, robotics, logistics, and intelligent devices. The next interface may not be a text prompt box, but a sensor, camera, robot, or machine tool that understands and acts within physical context.
This transition will trigger two distinct effects:
- First-order effect (automation): Tasks will become faster and cheaper.
- Second-order effect (redesign): Once AI can see patterns and adapt to context, the organization will change how work is structured. Maintenance will shift from fixed schedules to predictive models responding to actual machine behavior. Quality control will shift from catching defects post-production to identifying precursor process conditions.
Compounding advantages and challenges
Speed matters because the learning process will introduce compounding factors. Early pilots will face messy data, resistance, and failures, but each success will build valuable context, stronger data habits, and deep organizational trust.
Manufacturing leaders do not need to chase every AI trend. Rather, they should focus on practical inquiry:
- Where would operators benefit from guidance customized to their specific experience and tasks?
- Where would supervisors make better decisions by connecting quality, maintenance, and schedule data?
- Where are we forcing new technology into old, legacy workflows?
Ultimately, manufacturing businesses that stop at customization will secure better tools. The companies that go further will build better systems. Durable advantage will come from leaders who use operational understanding to fundamentally redesign how work happens around their company's unique capabilities.
