admin 
Ten years ago, I interviewed Dutch innovation consultant Gijs van Wulfen about his book “The Innovation Maze” on the AEC Business podcast. Gijs has spent 25 years structuring the front end of innovation using his FORTH method. He says technical feasibility stops only about 8% of innovations, while most fail due to a lack of trust and ownership.
When I invited Gijs back, I wanted to understand how AI is changing innovation. His short answer is that AI changes the speed, not the fundamentals. He compares the moment to the digitalization of the late 1990s, when everything suddenly had to be digital.
But are companies now trying to outsource innovation to AI?
Speed is the biggest change
Gijs describes innovators as generalists who connect technology, customer needs, finance, and strategy from problem-solution fit through launch. AI compresses that journey. A digital prototype that once took months or half a year can now be ready in days.
Even as technology evolves exponentially, the four starting points of the Innovation Maze remain valid: an idea, a technology, a customer problem, or a business challenge. Many companies today start with technology, driven by fear of missing out on AI. That is acceptable, Gijs says, as long as the technology addresses a relevant customer issue. Without one, you have a nice-to-have rather than a business model.
He also objects to calling AI a co-pilot. In his view, there is still only one pilot. AI makes that pilot smarter and quicker and helps avoid wrong turns.
A thousand ideas are no longer the problem
According to Gijs, AI democratizes imagination. In one FORTH project, the team generated more than 1,500 ideas, developed 18 concepts, and tested them with customers before building business cases. Generating ideas is now easy. Choosing the right ones remains the hard part.
AI can critique ideas and business cases, but it gives everyone the same middle-of-the-road answers to the same question. That does little to make any company distinctive, and AI cannot take ownership of the outcome.
“People trust people. People don’t trust machines,” Gijs says. If a team presents a business case and admits that AI made it, trust disappears. If the team uses AI to challenge assumptions and mitigate risks but owns the result, the board can believe in it.
Insource AI, do not outsource innovation
Gijs sees many AI-driven job cuts as window dressing for cost reduction. His advice is to make people smarter with AI first and redeploy them as processes become automated.
A Dutch installation company from one of his projects illustrates the point. The firm employs about 1,500 technicians who service heating, cooling, and water systems for large social housing organizations. With AI, a technician can identify a machine on site and immediately see its most common faults and what to check. Junior technicians without decades of experience benefit the most, and Gijs frames this as an effectiveness gain in a market that will have fewer workers anyway.
The same logic applies to startups and innovation hubs. European companies copy Silicon Valley’s hero-driven “I-novation,” but in Europe innovation is a collective effort, or “we-novation.” A startup’s solution is like the neighbors’ child: you like it, but you do not love it enough to feed it. “When the change outside your organization is faster than the change inside your organization, you will have a problem soon,” Gijs warns.
Innovators are risk reducers
Why is innovation so hard in large organizations? Gijs cites Peter Drucker, who noted that managers are hired to do things better, not differently. An idea carries 100% risk, while a board wants close to zero risk before saying yes. The corporate innovator’s real job is to reduce that risk step by step.
Managers typically innovate only when inaction poses a greater risk. In most markets, one or two companies innovate out of ambition, while the rest wait for necessity. Construction, like banking and healthcare, has good reasons to be conservative because buildings must not collapse.
Gijs has rarely seen truly ambidextrous organizations in practice. Instead, he takes innovation out of daily operations and treats it as a structured project, allowing the team to change its mindset and experiment on a small scale.
Where construction should start
His advice for AEC firms is to apply AI to internal processes first. Skilled craftsmen in their 50s and 60s are retiring, and they cannot simply be replaced. Firms should set ambitious targets, such as cutting construction time by 30 to 40% rather than 10%, reducing failure costs, and lowering energy consumption.
Because construction value chains are complex, companies need to innovate together with their suppliers in a structured way. Longer-term strategic questions include how to build more with fewer people and how to improve safety. I’d also add: how to ensure quality, because that’s one of the key challenges in today’s construction.
His closing message fits our industry well. Innovation is not about producing more ideas but about becoming more effective. AI can accelerate work, but ownership and trust must remain with people.
Learn more about Gijs van Wulfen at gijsvanwulfen.com and the FORTH method at forth-innovation.com. Listen to the full conversation on the AEC Business podcast.
