With a background in engineering and a strong focus on business strategy and creating customer value, Carl has built a career dedicated to developing products and scaling companies. His journey took a pivotal turn in 2016 when he entered the Architecture, Engineering, and Construction (AEC) industry, inspired by an architect passionate about addressing efficiency challenges through technology (aka Håvard Haukeland). This was the beginning of Spacemaker, a Norwegian company focused on transforming urban planning and design.
You probably know the rest of the story: together, Carl, Håvard, and Anders Kvåle turned Spacemaker into an international success, culminating in its acquisition by Autodesk. Today, Carl is the VP of Product for Autodesk Forma, continuing his mission to scale this innovative platform for a global market.
In this interview, Carl shares insights from his journey, discusses the evolution of AI, and reflects on his transformative experiences at Spacemaker.
Mari: You mentioned you’ve been building products to make other companies more efficient. How has AI helped in this process, or how are you applying this technology?
Carl: When we talk about digitization and digitalization, there’s an important distinction. Digitization is taking an existing process and making it digital. For example, you can take a paper form and make it a digital form, but you're not really reinventing or leveraging technology to do things differently—you’re just doing them the same way, but a little more efficiently.
Digitalization, on the other hand, is about leveraging technology to reimagine how you work. It’s about starting from scratch and seeing how technology can enable entirely new ways of doing things. I think AI is most useful when we take this digitalization approach.
Most companies aren’t ready for that, though. Working as a consultant, I often felt it was an uphill battle—getting results, but with a lot of friction in implementing those results.
So, let’s dive in AEC industry. In traditional building projects, things are so complex that you usually make one major decision at a time, like where to place the building and how big it should be. Then you figure out the other things, like the impact on neighbors, what materials to use, or the size of departments. It’s a sequential process because it’s too complex to consider everything at once. But ideally, you’d want to explore multiple alternatives and see the relationships and consequences, adjusting easily along the way.
So, when we started our own company, we really wanted to reinvent how to create buildings and cities by incorporating technology to make it much easier. We envisioned that AI could help users—architects or planning teams—who face complex decisions by providing more information on the consequences of their choices and the potential options they have. This enables them to work very differently.
Mari: How ready are companies to adopt this approach?
Carl: That’s what we set out to address with AI. My first interesting learning about AI implementation was around people’s perception of AI. People often view AI as a magical technology they don’t fully understand. Initially, things like Google search were considered AI, and technically, they still are, but people don’t think of it that way anymore because it’s so familiar.
When we approached customers, they were excited about AI that could “magically” generate building designs at the push of a button. They wanted to buy that, so we built it. But their teams struggled to use it meaningfully because the AI became a “black box.” They couldn’t work with the inputs or understand the outputs. One person compared it to only being able to use a driver on a golf course—standing far from the hole and having no way to adjust incrementally. Another compared it to having a self-driving car that goes where it thinks you want, without being able to tell it exactly where you want to go.
So, we shifted our approach to what we now call "AI on the shoulder," similar to what became Microsoft’s "co-pilot" concept some years later. The idea was to let the user feel in control, holding the pen, with AI as a tool that helps realize their intent. This approach makes the process more incremental, with smaller steps solving specific problems. It’s highly useful, but less “magical” or impressive on the surface.
For example, understanding the impact of a building in an urban area involves considering factors like noise, wind, and temperature changes between buildings. Traditionally, analyzing this is very complex, but we built a deep neural net that can simulate and predict this instantly. Designers can work live with their buildings and understand consequences right away, allowing them to make adjustments in real time. While groundbreaking, it doesn’t feel like “AI magic” to users—it’s just information readily available to them.
To me, this has been a key to success: to incorporate AI into people’s processes so naturally that they adopt it without even thinking about it as AI.
Mari: In your experience, one of the challenges seems to be that companies and professionals aren’t fully prepared to explore all that AI offers. How do you think companies can better prepare for AI?
Carl: I think a simple answer is to treat this as a type of change management. It’s about fostering a productive sense of urgency. I’ve seen teams and customers respond in two unproductive ways: either they’re complacent, not thinking AI is a big deal for them, or they’re anxious, fearing AI will take over jobs or disrupt their income and business models. This anxiety makes them reactive, while successful customers are proactive, actively engaging with AI as an opportunity.
Leadership plays a key role here by creating a culture that understands AI as existential. When new technology like AI arrives, it won’t replace people who are the best users of it, but it will give those users a competitive edge. Over time, I’ve seen this pressure lead to more companies understanding that it’s less about technology itself and more about the right attitude.
Mari: How do you compare Norwegian startups in industrial software to American startups?
Carl: In Norway, particularly in B2B and industrial sectors, customers are often very digitally advanced and open to change, sometimes more so than other places. They tend to be trusting and willing to try new things, even with smaller companies that might not have the robustness of large firms. There's generally low marker barriers to start in Norway, with customers open to experimentation and sharing information about their processes.
In the U.S., however, there’s more competition and a higher expectation for readiness—customers don’t want to waste time. It’s a much larger market, so while American customers are usually easy to work with, it’s harder to differentiate because of the increased noise. In Norway, the smaller market makes it somewhat easier to establish a network and build proof points.
Mari: What should we expect from this technology in the future?
Carl: That’s a great question. We could view the latest generation of generative AI as a new technological revolution, though it’s typical for people to get overenthusiastic, expecting everything to happen quickly, only to be disappointed. But then, eventually, it does happen—it just takes longer than expected. Look at the internet: concepts from the early 2000s eventually became reality, but some took 10 years, not two.
I’m confident generative AI will bring significant changes. It’s realistic to expect robots doing most physical work, and AI handling much more cognitive work. This will fundamentally change our reality, probably sooner than we think. The best way to prepare is by having a constant sense of urgency, viewing this not with anxiety but with excitement about using technology to achieve more.
On a company level, it’s about systematically evaluating where money is spent and how to get the same results more efficiently. On an individual level, it’s incredibly empowering to look at AI as a tool to get more done with less effort. If you’re proactive and focused on doing more of the exciting, valuable work, AI can be a powerful ally.
Mari: Would you say you’re optimistic about what lies ahead in the next five years?
Carl: Absolutely. Being at the forefront of technology is thrilling, especially as the interface with technology has become more democratized with large language models. Five years ago, machine learning was seen as inaccessible, something only “the machine learning guy” would handle. But now, if you can talk or write, you have everything you need to interact with technology effectively.
This shift is part of the attitude change I mentioned. Excuses like “I don’t understand technology” don’t hold anymore. It’s accessible to everyone, and we should take full advantage of that.
“The best way to prepare is by having a constant sense of urgency, viewing this not with anxiety but with excitement about using technology to achieve more.”
— Carl Christensen
