I’m seeing two very different strategies as engineering/design/manufacturing/construction companies think about artificial intelligence: On the one hand, many bigger companies have teams (usually in IT) who stay on top of machine learning, large language models, and the rest of the big bucket of technologies that make up “AI.” They’re thinking about applications they can build using today’s AI toolkits to create a competitive advantage in a specific part of the business. AI agents (aka agentic AI) are already responding to many HR, IT, and sales inquiries, predicting demand to tune supply chains, and doing other tasks with clearly defined rule sets. They haven’t made it into pure engineering/design and manufacturing yet; we mostly see them in the business-adjacent functions, like tracing requirements, vetting components and suppliers, and tracking deliverable status.

Many companies, though, are in serious denial: they don’t think the technology is ready, they don’t think it will apply to them and their business needs — or they simply don’t have the time to think about it. The progressive ones in this cohort think, “We’ll let our vendors figure it out, sell it to us, and then we’ll be “AI,” too.” That’s better than ignoring it altogether, but not sufficient to be successful in the long run. If you’re buying a solution from a vendor, your competitors are, too — your advantage will come from how you use it, so start thinking about this now.

At a human level, though, we are massively unready to deal with the changes AI will bring to almost everyone’s job — mine, yours, your lawyer’s, and your accountant’s. Perhaps not the craftspeople on the shop floor or construction site since they still need to install that switch. However, AI could change where they get the installation instructions and how they show that they’ve tested it. AI will help us know what task to do first, provide us with the information we need to do that job, organize what we create, and report that we’ve successfully done today’s tasks.

When will AI affect me or you? That’s the many, many dollar question. One person I spoke with said that some entry-level jobs (customer-facing support, marketing tasks like building prospect lists for sales campaigns, some coding) may disappear within the next 12 months. Disappear, as in, not coming back. Others don’t see that quick a timeline, partly because of concerns that AI might serve out incorrect or incomplete information — the trust hasn’t been built yet.

But sooner or later, agentic AI will be here. If you listen to its developers, it will be faster and cheaper than human workers and won’t require meal breaks or days off. It will learn from its mistakes and know when to hand off to another agent because it’s been asked to do something outside its scope. We’ll have to see about exactly how far, how quickly commercial AI gets.

What can or should we do to get ready? I’ve collected a couple of ideas as I speak with people about this:

1. Don’t ignore it. Putting our heads in the sand may be comforting for now, but it won’t help you individually or your enterprise as your competitors race toward AI.
2. Learn everything you can about AI. There are many sources of information online, including classes from Google, Microsoft, Amazon, OpenAI, Coursera, MIT, and others. (Make sure you verify the creator; there’s also a lot of misinformation online.) Unless it interests you, focus on applications of AI, not the nuts and bolts of how it works — you’re redefining your job, not designing the next large language model.
3. Dedicate part of every day or week to figuring out what AI (agentic or not) can do to help you do your job better. Early successes show that modest, targeted solutions improve productivity; trying to go too big takes longer and is not as assured of success.
4. Focus on real-world business value, not the gee-whizz-ness of the technology. The tech by itself isn’t the answer; we need to rethink how we do what we do, taking advantage of this new technology. An analogy is word processing: we went from secretaries typing everything to having the creators writing and typing simultaneously. AI is that kind of change by times a bazillion. 
5. If you are a boss, look at your team. Who is most receptive to new technologies? Are they experienced at their job? If not, pair a tech-savvy new employee with an experienced one; we need to know how a task is done and why it’s done the way it is before we can change it to take advantage of AI.
6. Recognize that not everyone will be on board with this whole AI thing. Try to help people understand that their jobs will change and encourage them to help define the new normal. Be realistic about how your workforce might change, given tech and economic upheaval.
7. Work with your vendor ecosystem. They are ALL working on how they can add AI to their solutions, so define what you need and tell them! In turn, you’ll benefit from their broader view across many customer companies.

Bottom line: Assume you’ll be the boss of a bunch of AI agents by this time next year. Or the year after; who knows? What do you want them to do? How can they assist you? If it helps, think of them as an infinite number of really smart interns who have no HR issues and never need to rest. How can they make you more successful? What do you have to teach them to get to that point? What’s the best possible outcome when you’re paired with those agents/interns?

For example: Do you design widgets? Can AI help you find old designs to repurpose, speed up your simulation program, discover new parts or vendors to ease supply chain issues, or track compliance to ensure that you meet marketing’s design objectives? Can it automatically generate drawings or other deliverables (to what specs, using what as inputs, with what guardrails)? At a minimum, learn to work with the AI assists/co-pilots built into the PLMish products you use to do your job more effectively.

Start thinking about all of this now so that you’re ready when someone asks you how you see AI reshaping your job. “We want to use AI for business benefit” isn’t enough; define a specific problem you want to solve and metrics to tell if you’re getting closer to that goal. Go!

—

That gorgeous cover image is from NASA and has absolutely nothing to do with AI. But it’s a small, perfect planet, no?