Building Relationships with Technology

Treating technology as a human replacement strategy is a recipe for failure. Here’s the questions you should be asking.
by Sean Patterson

I have good news: Your board just approved hiring a chief of staff for every single employee in your organization.

This person never sleeps. They learn instantly from every interaction and get better with use. They have encyclopedic knowledge across all human domains, from circuit design to supply chain optimization to regulatory compliance. And the cost? As little as $20 per month per employee.

There’s just one catch: They succeed only if you become their coach and manager.

I’m talking about AI, but not the way you’ve been hearing about it. This isn’t another conversation about implementing AI technology or buying AI software. This is about building relationships with technology. It’s about transforming your operations by developing AI capabilities with your whole team, not deploying solutions onto your team.

Most importantly, AI adoption cannot be delegated. It’s a personal journey for every person in your organization – yourself included. Especially you, the reader of this article. Regardless of position.

If that makes you uncomfortable, good. It should. Because after 25 years leading operations across submarines, PCB manufacturing, Amazon’s logistics expansion, healthcare scale-ups, 3-D printing and startups, I’ve learned one thing: every time we treat technology as a human replacement strategy, we fail. Every. Single. Time.

The companies winning with AI right now aren’t asking “How do we deploy AI?” They’re asking, “How do we teach our teams to develop AI?” That’s a massive difference. And it changes everything about how you need to approach what’s coming.

Why Technology Investments Fail (and AI Could Too)

Here are the hard data no one wants to talk about: despite trillions invested in technology since 2005, US productivity growth collapsed to half its historical rate – from 2.3% annually to just 1.3% – costing workers tens of thousands of dollars in foregone wage gains.¹ Smartphones and social media – technologies we thought would revolutionize work – actually correlate with productivity decline, not improvement.

History offers a sobering lesson. When electrification came to manufacturing, it took 40 years to deliver productivity gains. Why? Because it required complete organizational restructuring and workforce retraining.² Companies that just installed electric motors without rethinking their workflows saw minimal improvement. The technology alone wasn’t enough.

AI has projected economic impact of $13 trillion – $15.7 trillion by 2030, according to McKinsey and PwC.³ But that will happen only if we learn from history and prioritize human capability development over software accumulation. Technology combined with training delivers productivity increases, not the technology alone.

Here’s what frustrates me most: people jumping on the AI bandwagon because their board told them to. They hear the media hype and say, “go put AI into our company” or ‘buy an AI automation.” That’s exactly backward.

In any other part of your business, when would you tell people to do something you don’t understand as a leader? We would never do that. If you’re a PCB manufacturer and someone said, “I hear there’s a new trend in making clothing with advanced looms – go figure out how to do that,” you’d say no. It’s not your business. You don’t understand it. So why do that with AI?

The companies that win with AI don’t replace, they amplify. And they start by teaching their leadership and teams to understand what they hear they need to implement.

Leading from the Front (and What that Actually Means)

AI adoption is not a grassroots campaign you can delegate to your engineering team. It’s a take-the-hill campaign, and the technology is changing too fast and is too impactful to lead from the sidelines.

Leading from the front means learning new skills yourself, no matter what level you’re at in the organization. If you’re not becoming a lifelong learner around this technology, you’ll be left behind. You can’t just say “I learned it” and be done. This is the difference between buying a gym membership and actually getting fit. Many people make the New Year’s resolutions and go to the gym on January 1. They learn the equipment. But showing up every day for the next 365 days to make it a habit? That’s the hard part. Learning when your muscles are sore, when to change your routine, and in the AI gym, there is new equipment showing up every day. The only way to know if it’s the next fitness craze or one that makes an impact for you is to try it. Everyone’s body is different. And the very language we use with AI affects the outcomes. You can’t copy your way to proficiency.

AI is the same way. It needs to become a habit. Being “AI-first” doesn’t mean your company becomes some artificial network in the sky. It simply means that upfront, you deliberately choose whether to partner with AI before starting each task. If you don’t do that, you’re not an optimized company, and you will fall behind.

Here’s the mistake I see constantly: Even when people learn AI, they think about using it retrospectively – after they’ve already done a lot of human work. The problem? You’ve already wasted time. Worse, you’ve sent AI down a solution path pre-constrained by your mental model and life experiences. You didn’t take advantage of AI’s encyclopedic knowledge across multiple disciplines.

AI can help you work on the problem first, then determine solutions to implement. But it’s a good soldier – if you tell it to implement the solution you want, it will. And that solution is only formed from your experiences. That’s not the AI advantage. That’s where you go to Google.

Stop Googling your interactions with AI. AI is a thought partner, not a search engine.

And here’s why you can’t delegate this to just a few people: if a company only invests in some employees to adopt AI while others are written off, you create a two-tier workforce and a massive change management disaster. Those left behind will fear for their jobs – and that fear becomes reality when you’ve caused the problem. But if the majority of your company adopts AI, you’ll scale your company without growing headcount.

This is especially true today, as lean startups are forming with highly AI-leveraged teams. Most reading this article are legacy manufacturers. When you implement AI, you need to do it across the board to be as lean as those startups – and you need to do it before they scale up and take market share.

The 3 Pillars of AI Maturity (and Why You Can’t Skip Steps)

Manufacturing executives want to jump straight to automating processes, measuring ROI and calling it done. That looks obvious on paper. But this approach fails nine times out of 10.

Why? Because you’re asking strangers to solve problems they don’t understand, using technology your team can’t maintain and creating solutions nobody trusts. It’s like hiring someone to install an ERP system. How’s that working out for you years later? Do you get the value from it? Research shows 55-75% of ERP implementations fail, and 80% of customers are unhappy with their current ERP.⁴ I suspect you have more people to maintain it, use a quarter of the features, and you’re still late to your customers and don’t know how to price better than throwing a dart at a dart board.

Here’s the fundamental difference: when you outsource AI, you get automation. When you build AI capability internally, you get intelligence.

Based on coaching over 1,000 people across manufacturing in the past six months, I’ve identified three pillars that prevent AI adoption failures. They must be completed in this order. They are not phases; they are foundations.

Pillar 1: AI for Everyone. This is about making every engineer, operator and manager more effective at their current job. Not replacing – amplifying. Start simple: using ChatGPT, Claude or similar chat systems. What matters is that you start using an AI chat system, period. The tool choice is reversible. The habit formation isn’t.

You need to spend several weeks here until using AI becomes your first thought, not your afterthought. Learn how to go from simple to complex questions. Learn how to extract what you want in the format you want. Even as you advance to Pillars 2 and 3, you’ll still interact with those systems through Pillar 1 methodology; i.e., you will chat to create Pillar 2 and 3. This is your foundation.

In Pillar 1 you learn things like:

• Humans are horrible about assuming context and not providing enough.
• How to prevent hallucinations (when AI “makes” things up), not to use AI for out-of-the-box facts, because – surprise – AI doesn’t know facts. It probably knows facts, on a statical basis.
• On the other spectrum, you’ll learn what too much context is.
• How LLMs (e.g., ChatGPT) should not be used for deterministic outcomes (i.e., anything mathematical).
• How your language and word choice dramatically affects AI responses. Precision in questions yields precision in answers.
• When to iterate vs. when to start fresh. Sometimes the conversation needs to be abandoned and restarted.
• How to spot AI’s confident-sounding mistakes and develop your verification instincts.
• Which tasks AI handles brilliantly (pattern recognition, data synthesis, drafting) vs. poorly (final decisions, creative judgment calls).
• That the best results come from treating AI like a new hire: you teach it your world, your constraints, your quality standards over multiple conversations.
• When you get into “real AI” you’ll have to pick models. Don’t just fall back to generic ChatGPT that tries to be everything to everyone. It’s great and easy to use, but not for your business outcomes, and every competitor can use the same thing!
• Security practices.
• And more.

Here’s a framework I’ve developed through working with hundreds of manufacturing professionals. I call it CRIT (Context, Role, Interview, Task):

Context: Give AI your manufacturing environment. What line you’re running, what customer specs you’re meeting, what constraints you’re working within.

Role: Tell AI what expert perspective you need. Process engineer, quality manager, maintenance tech. AI needs to know whose brain you want it to emulate.

Interview: This is the key. “Tell AI to ask YOU questions. Have it provide more context to better accomplish the task. One question at a time.”

Task: Give AI a specific, actionable output in the format you want.

Bad question: “Why am I getting registration issues?”

Good CRIT problem statement:

Context: I’m working with an 8-layer HDI board using EMC EM-888K halogen-free prepreg and EM-827 core. Customer requires Class 3 registration tolerance, but I’m seeing 4-6 mil drift on layers 3-6 after lamination. (*Recommend uploading the lamination section of the PDF version of the HDI Handbook as well.)

Role: You’re an expert PCB process engineer specializing in multilayer lamination.

Interview: BEFORE starting the Task, ask me questions, ONE AT A TIME, to better understand the context and task I’m asking about.

Task: Give me three specific lamination trials to run this week, ranked by probability of impact, with target press cycle parameter changes and measurements to collect.

See the difference? You’re not querying a database. You’re working with an informed expert.

Note: the interview process is the critical differentiator. When AI asks you questions before solving your problem, it’s gathering the context that makes its recommendations specific to your situation, not generic internet advice. This is where AI stops being a search engine and becomes a thought partner. Most people skip this step and wonder why they get mediocre results.

Pillar 2: AI in Business Workflows. Once your team is comfortable coaching AI for individual tasks, Pillar 2 is about automating the repetitive ones. Here’s what the software vendors don’t want you to know: you don’t buy Pillar 2, you build it yourself using the Pillar 1 tools you just mastered.

Most of what’s being sold as “AI software” for manufacturing can be built by anyone – no coding required, just the ChatGPT conversation skills you developed in Pillar 1. If you don’t reach Pillar 2, you’ll remain stuck copying and pasting into a chat window instead of having AI work in the background while you focus on higher-value problems.

Here’s the key insight from teaching over 1,000 people in the past six months: everyone, regardless of background, age, or technical experience, is capable of creating automations. And those automations will be built by the very person who understands the problem!

Think about it: that quality engineer who’s been manually copying data from inspection reports into Excel every morning? They’re the perfect person to build the automation that eliminates that task. They know exactly what data matters, which edge cases to handle, and what format actually helps decision-making. And then they get time to get on the floor and actually solve the problem and implement the solution. Last I checked, Excel doesn’t solve quality problems; too often, we’ve relegated ourselves to data gathering rather than solution-solving with data insights. Too much of the job today is spent on the non-value-add portion of that process.

A vendor or IT department guessing at their needs will build something nobody uses. But when that employee builds it themselves using AI assistance? They’ll use it every single day, because it solves their actual problem in their actual workflow.

We will cover this in more depth in future articles.

Pillar 3: AI for Competitive Advantage. This is where you connect AI directly to your manufacturing systems and create intelligence that lives inside your ecosystem. Competitors can copy your processes, poach your people, buy the same equipment. But they can’t copy AI trained on your specific customer requirements, supplier variations and quality challenges. This is the era of bespoke software – that can connect anything, even that 30-year-old DES line at a fraction of the price.

This is where AI stops being a productivity tool and becomes a strategic moat. Imagine AI that knows your plating chemistry better than your most experienced technician because it’s analyzed 10 years of bath data and correlated them with defect patterns. Or AI that predicts which customer orders will have yield issues based on design characteristics only your facility has seen. That’s not something you can buy off the shelf. That’s intelligence built from your data, your expertise, your unique challenges. And once you have it, you’re operating at a level your competitors simply cannot match without going through the same journey.

Again, this will be covered in future articles.

But remember, skip Pillar 1 and jump to buying automation? You’ll never scale real benefits. You’ll get expensive software nobody uses instead of an intelligent transformation.

What You Do Tomorrow Morning

Don’t wait for permission. Don’t wait for the perfect plan. Don’t wait for your competitors to figure this out first.

Tomorrow morning, open ChatGPT or Claude (my personal favorite for businesses). Pick one problem that’s been bugging you for months – a quality issue, a process bottleneck, a documentation nightmare, a customer communication challenge. Anything.

Spend 30 minutes teaching AI about your specific challenge using the CRIT framework:

• Give it context about your manufacturing environment
• Assign it a role as the expert you need
• Tell it to interview you
• Specify the task and output format you want.

That’s it. That’s how the journey starts.

Not with a vendor contract. Not with a consultant engagement. Not with a board presentation. With 30 minutes and one problem you actually care about solving.

Your AI teammate is ready. They don’t need recruiting, onboarding, or vacation time. They’re waiting for you to teach them about your world, your challenges, your expertise. The only question is: are you ready to stop deploying technology and start developing your team’s capabilities? Because that’s what separates the companies that will thrive in the next decade from the ones that will wonder what happened.

About This Series

This is the first article in a series exploring practical AI adoption for PCB design and manufacturing professionals. Future articles will dive deeper into the Three Pillars framework, showing you how to:

• Build workflow automations without coding (Pillar 2)
• Create competitive advantages through AI-powered systems (Pillar 3)
• Capture veteran knowledge before it retires
• Implement DIY equipment monitoring for legacy systems
• Use AI for supply chain intelligence and collaboration.

Each article includes concrete prompts, real manufacturing examples, and actionable Monday-morning takeaways. This isn’t about buying AI solutions. It’s about developing your team’s AI capabilities.

References
1. US Bureau of Labor Statistics, “The U.S. Productivity Slowdown: An Economy-Wide and Industry-Level Analysis,” Monthly Labor Review, 2021.
2. NBER Working Paper 28076, “Powering Up Productivity: The Effects of Electrification on U.S. Manufacturing.”
3. McKinsey Global Institute ($13 trillion estimate) and PwC ($15.7 trillion estimate representing 14% global GDP increase by 2030).
4. Gartner Research on ERP Implementation Outcomes, 2024.

Sean Patterson is an accomplished executive with extensive C-suite experience across CRO, COO, and CTO roles who now specializes in humanizing artificial intelligence implementation in business environments, particularly manufacturing; sean@crossgen-ai.com. He will keynote the PCB East management session on April 28, 2026.

Patterson’s unique approach to AI implementation stems from his multifaceted leadership experience in the PCB industry, including serving as COO and CTO & head of AI at Summit Interconnect, various senior positions at TTM Technologies, and CRO of Nano Dimension. He built Amazon’s tractor trailer division and healthcare platforms. He currently serves as COO of StartGuides, providing military technology working backwards from the soldier. He is also on several nonprofit AI advisory boards in education.

Patteson brings practical insights into how PCB manufacturers can approach AI adoption strategically. His methodology emphasizes cultural adoption from the top, employee empowerment, and then automation. His approach to AI implementation is captured in his often-quoted principle: “AI adoption is not something a leader can delegate.”

Patterson holds a master’s in nuclear science and engineering from MIT and a bachelor’s in systems engineering with a focus on robotics from the United States Naval Academy.