Key Takeaways:
- “I don’t know where to start” is the most common AI adoption blocker, and it masks a deeper fear of looking incompetent in a domain where competence was always the source of professional respect.
- Perfectionism and committee-building are forms of delay dressed up as diligence, and the people closest to the work already know where the friction lives.
- The leaders who move first are willing to look like beginners in front of their teams, which takes more professional courage than having all the answers.
- One workflow improved this week breaks more paralysis than six months of strategy development.
A CFO of a $400 million company told a colleague of mine, “I know AI matters. My board knows it matters. But I don’t know where to start, and I don’t have time to figure it out.”
That statement captures the most common belief blocking AI adoption in engineering and technical firms, and I hear versions of it every week. From principals running 20-person EPC firms, from engineering managers at industrial distributors, and from founders who built successful businesses on technical expertise and now feel paralyzed by a technology domain they did not grow up in.
After implementing AI across 40+ engineering projects, I can say with confidence that “I don’t know where to start” is almost never a logistics problem. It is three different fears wearing the same mask.
Fear #1: Status Protection
You built your career by knowing the answers. For 10, 15, or 25 years, your value was rooted in expertise: calculations, codes, process knowledge, and project judgment. Now there is a domain where you genuinely do not have the answers, and asking for help feels like admitting you fell behind.
In engineering culture, where technical competence is the currency of respect, that admission carries real weight. But the leaders who move first are the ones willing to look like beginners in front of their teams, and in engineering terms, that is the equivalent of calling for a peer review instead of signing off on a design you are uncertain about. It takes more professional courage to ask for input than to pretend you have it figured out.
Fear #2: Perfectionism Disguised as Diligence
The second face of this paralysis is the belief that you need a comprehensive AI strategy before you take the first step. So you form a committee, attend conferences, evaluate platforms, and build a business case while months pass without a single workflow changing.
Meanwhile, Section’s January 2026 AI Proficiency Report (which assessed over 5,000 knowledge workers) found a striking perception gap: only 5% of individual contributors say AI has had a transformative impact on their work, compared to 42% of C-suite executives. The gap between executive AI confidence and ground-level reality is enormous, and no amount of strategic planning from the top will close it without actual workflow change at the front lines.
LinkedIn co-founder Reid Hoffman put it directly: AI lives at the workflow level, and the people closest to the work know where the friction actually is. They do not need a strategy document. They need someone to say “go ahead and try it on this one task.”
Fear #3: Waiting for Safety
The third face is a quieter belief: that if you wait long enough, the right answer will become obvious and the risk will disappear.
The tools actually get cheaper and more forgiving every month, but the muscle of daily AI use takes real time to develop. Those two facts move in opposite directions, which means the cost of waiting is not that you will pay more for the technology. The cost is that your competitors who are building AI fluency now will compound their advantage quarter over quarter, and the firms that wait will face the same learning curve later with less runway and more competitive pressure.
I have seen this pattern clearly across the firms we work with: the ones who started six months earlier are not just further ahead on tools. They have internalized a fundamentally different relationship with AI, where experimentation feels normal instead of risky. That cultural shift cannot be compressed, only started sooner.
What the Paralysis Actually Costs
In my practice, we have seen the math clearly. An engineer billing at $150 per hour who spends 5 hours per week on tasks AI could handle generates $750 per week in opportunity cost. For a 5-person team, that is $3,750 per week, or roughly $180,000 annually based on 48 working weeks.
Those are hours our clients have measured and reclaimed in areas like drafting proposals, reviewing contracts, generating documentation, formatting reports, and writing client communications. These are tasks that matter but do not require 15 years of engineering judgment to execute well.
The Antidote: One Workflow, This Week
Two and a half years ago, I faced an impossible deadline with no budget for a specialist. I used AI to complete expert-level calculations in 4 hours instead of waiting 3 weeks for a consultant. The project stayed on budget, the client stayed happy, and I learned something critical: the first step does not need to be strategic. It needs to be specific.
In our Crawl-Walk-Run-Sprint methodology, we tell clients to start with the task they resent doing most. Not the most important or the most complex, but the one that drains their energy every week because it is repetitive, time-consuming, and does not require their best engineering judgment.
For most firms, that first Crawl phase looks like meeting summaries, email drafting, internal documentation, or formatting deliverables. Nobody gets fired for making those faster, and the confidence that builds from early wins creates momentum for bigger implementations.
One participant in our Fall 2025 AI Accelerator, John Fasano, described the program as introducing “new possibilities and new ways to think about old problems.” His sales team went from dreading routine customer check-in emails to automating the scheduling, drafting, and sending of those messages, freeing time for actual relationship building.
That was a belief transformation, and the technology followed once the belief changed.
The Permission You Are Looking For
You do not need to understand all of AI, and you do not need a comprehensive strategy or a committee. You need to give yourself permission to be a beginner in one small area, apply AI to one specific workflow, and measure what happens over a few weeks.
The free AI ROI Calculator walks you through the exact hours and revenue your firm could recover by deploying AI on the work that’s quietly draining your engineers.