AI and the Future of the Trades: A Machining Instructor’s Case for Human Hands

Adam Hathaway teaches Computer Numerical Control machining at Chabot College in Hayward, California. His shorthand for the work: "We make metal smaller." He grades student code line by line, because a moved decimal point can crash a machine worth thousands of dollars, so his grading was slow, unforgiving, and eating his time.

That's what sent him to AI in the first place. He built himself a Nectir AI Course Assistant grounded entirely in his own materials, down to the cutting-speed charts and equations he teaches, and it now reviews thousands of lines of student code on his rules. You can read the full story of how he built it in Instructor-Controlled AI in the Machine Shop.

But building the tool gave Adam a bigger opinion about where AI is taking his field and what teaching looks like on the other side of it.

AI needs human hands

At the start of one fall semester, a student came into Adam's shop rattled. He'd read an article naming CNC programming as one of the first jobs AI would eat.

"Whenever there's a change, I think we all get a little paranoid, because the potential is so new, and it hasn't settled into that position yet in our society."

Then it settles. And when it does, Adam says two things tend to be true. It's rarely as bad as the early headlines. And it usually opens up secondary and tertiary markets that turn into real jobs. Machining has been through this before, which is why he wasn't worried for his students, and isn't worried now.

Machining, Adam argues, is oddly hardened against the AI panic because a form of AI has lived inside the work for years. Machinists already use software that can look at a 3D model and propose its own tool paths. Ask anyone who's run it, and you'll get the same verdict.

"It is mostly good. Which means that it is partially wrong, and that it requires a knowledgeable human to observe and correct those problems when they come up."

Somebody still has to stand at the machine, change the tools, load the parts, and catch what the software gets wrong. That's the job, and it isn't going anywhere.

"We've already been doing this for a while now. It's got problems. And ultimately, you still need a person at the helm of this thing at the end of the day."

The AI hype doesn't rattle him because it describes something he's already lived. He got an email promising AI would take 80% of the programming time out of a part, which, he notes, is about what he's already seen from other software over the years. He's watched that number climb with every new tool. The work didn't disappear any of those times.

Adam points out that decades of automation in his field kept producing higher-skilled, higher-paying work, and that machining right now is a hungry job market, growing partly because the AI boom itself needs chips, and chips need machinists.

What he wishes more faculty understood

Adam is sympathetic to nervous colleagues. AI showed up in education "as more of a boogeyman than an assistant," and he understands the discomfort.

That discomfort is often a signal worth listening to. If a course falls apart the moment students have AI, or if you can't tell whether a paper is theirs, the fix might be to look hard at the assignment itself. Rebuilding a curriculum takes time and a lot of work, but it could be the best next step for education.

He frames it as a change most educators can't sit out. AI is part of the reality of the classroom now, and the useful question isn't whether it exists; it's how you teach with it in the room.

"I think we need to start rethinking what our roles are as educators, and not ignoring the fact that AI exists, but with AI existing, how do we proceed?"

Adam’s own program has an advantage here, and it points to where he thinks teaching is headed. He sees it as the same shift already gathering steam under the name project-based learning. A machining project is a recitation of knowledge that only works if the student can actually apply the knowledge to get a real result. You can't fake your way to a part that runs.

"It's not just to get answer A or answer B. It's to understand knowledge and apply that knowledge, and AI provides a lot of resources for that sort of learning."

The reason he trusts the tool at all comes down to control: he owns the knowledge, the prompt, and the output. Read more about how Adam built an AI he could trust.

FAQ

What is Nectir AI? Nectir AI gives colleges, universities, and high schools the ability to deploy AI Assistants across their campus that are fully controlled by faculty and administrators, built into existing learning management systems, and compliant with FERPA and SOC 2 Type II standards. Nectir is trusted by more than 130 schools and 150,000 students and educators worldwide, including a partnership with the California Community Colleges, which serves 2.2 million students across 116 campuses.

Can AI actually help in hands-on, technical courses? Yes, because faculty control the guardrails and the Assistant draws only on its uploaded materials, it can guide a student toward applying knowledge to a real result instead of handing over an answer. In Adam's shop, it checks student code against the correct master and flags the exact line that's off.

How can I learn more about Nectir? Want to see what AI support looks like when it's built into the tools your campus already uses and controlled by your faculty? Schedule a demo, and our team will walk you through how it works at schools like yours.