The idea of Artificial Intelligence is not new, but it was in the realm of science fiction, set in the distant future. The future came sooner than expected. In the past year, with the launch of applications, like ChatGPT, AI has leapt from science fiction to science fact. Now widely available, it already has become a game changer and disruptor in many industries. It inspires excitement and fear and, in the workplace, results have often been mixed. But what is the current reality for this technology in architecture and construction?
“We are excited by the promise of AI and believe the creative world, including architects, can truly benefit from its access to a wide range of data, its analysis and subsequent optimization of that data, as well as its generative abilities,” says Matthias Hollwich, founding principal of HWKN Architecture. “It has supercharged how we work as a firm, impacting how we plan and experience buildings and cities. We have developed a design process that vastly improves not just outcomes for our clients but for the communities where we design. AI allows for inputs from a much wider range of stakeholders. We think when communities and others in our profession see its benefits, it will alleviate some of the concerns surrounding its use.”
“Construction was one of the early adopters of AI,” says Katie King, CEO of the United Kingdom-based consultancy AI in Business. “There have been some great use cases in recent years, and I would say the biggest gains have been made in efficiency, safety and design. Some businesses have begun using generative AI for construction design. This enables construction designers and engineers to optimize and test different versions of their designs with less effort, less speculation and without having to invest resources into trying out different options.”
“Currently, I see AI focused primarily on visualization,” says Michael Schroeder, chief technology officer and partner at architecture firm SGA. “For example, you could take a traditional architectural napkin sketch,
bring that into a diffusion model, and say, ‘let’s give this environmental qualities or material qualities.’ AI can build you a photorealistic rendering quickly. There will be many errors, of course, because you can’t napkin sketch to scale. But it’s amazing how these tools can generate something that used to take days of rendering.”
“Text-to-image generators seem to have most of the attention among architects now; however, many tools are relevant to the architectural process, and new ones emerge every few months,” says Damon Leverett, AIA, senior lecturer, University of Arizona School of Architecture and the College of Information. “Additionally, we asked students to develop conceptual structural sizing with AI chatbots and compare their results with traditional manual calculations as a critical-thinking exercise. The AI toolkit also includes several architecture-oriented technologies for design and modeling, sketch recognition, energy efficiency and sustainability analysis, project planning, plugins for CAD/BIM and tools to foster compliance with building codes.”
AI is being used to speed up or enhance many tasks with new and creative ways to utilize the technology being rolled out every day. Artificial Intelligence is poised to become part of doing business, not just for architects, but for everyone in construction.
“Other uses for AI include scheduling, document management, equipment maintenance, predictive modeling and performance management,” King says. “I have also heard of firms using AI for offsite construction, using automated manufacturing to assemble different elements, such as walls, HVAC and other components that can be transported to the job site.”
The Role of AI
There are legitimate concerns about potential negative impacts of Artificial Intelligence. The technology’s rapid growth inspires anxiety, and many people worry about AI replacing jobs traditionally done by humans. There are different points of view on this issue, but it appears much of what AI brings to construction today is more to supplement rather than replace the work of human beings.
“The intent is not to introduce technology to supplant traditional methods but to examine and compare its outcomes with the natural world,” explains Leverett. “My generation was a part of the transition from hand drafting to CAD in the mid-1980s. I see the same questions and concerns architects expressed then, mainly, ‘Will AI replace architects or cause me to lose my job?’ Just replace the word AI with CAD, and it becomes apparent that the concerns are similar. History tells us that disruptive new technologies cause a shift in labor but usually result in more jobs, not less, but the new jobs are different. CAD was introduced as a significant evolution of the design process, and we are still here.”
“While AI is still new and intimidating to some, the creative world can truly benefit from the data, analysis and optimization it brings,” Hollwich explains. “AI is poised to revolutionize how we plan, experience and manage buildings in the near future, ultimately transforming the architectural landscape. Many designers are embracing the technological evolution of these tools and I’m excited to see what the future holds for architecture, enhanced by AI.”
With the power and hype around AI, it’s easy to think it is an all-powerful panacea or not of any real use. Reality lies squarely in between, and proper application of AI means understanding its strengths and limitations. It also may require an all-new way of thinking about the interaction between people and technology.
“AI is not infallible, and many of us fall into a trap of trusting it too much,” King explains. “It’s smart but not perfect. It lacks the contextual understanding an judgment that humans possess. We need humans in the mix to apply judgment to AI outputs. If there’s no oversight, things slip through the cracks and mistakes happen. In construction, those mistakes come with high stakes. Inaccuracies and cut corners can lead to safety issues, structural issues or suboptimal design, so AI adoption needs to be a partnership rather than a takeover.”
“I don’t see AI as a tool as much as I see it as another member of the team,” Schroeder says. “A screwdriver is a tool, but AI is quite a bit different. With AI, you have another member of the team walking on a job site. At our firm, we have a people-process-technology approach. We’ve had to expand that. AI is not a person. It’s not a technology, per se. It’s not just a process, but it is going to change how we do things. I think at this point, AI needs its own little bucket. We have to think about it differently.”
Safety and Policy
Given how rapidly the power and utilization of AI is growing, it raises many alarms about safety and security. Careful thought and policy action needs to be taken to ensure it is used properly. Perhaps more than any technology before it, AI requires firms and individuals to be very clear about what they want and do not want it to do.
“We recognized that AI use in consulting has significant opportunities and risks. This is new and uncharted territory,” explains Brie Zoller, chief problem solver at Brie Z Operations, a small business operations consulting firm. “Our main concern is that we have a duty to protect our client’s data and information. AI is so powerful and uses the data we provide to evolve. We need to be careful about what information we are entering into it.”
In developing policy around AI use for her client, Lotus Sustainability & Engineering, Zoller has worked to address security gaps and anticipate potential uses and pitfalls. Zoller suggests that firms revisit their AI policy often—perhaps even quarterly—simply because the technological landscape is changing so quickly.
“We created a list of approved uses for AI, such as idea generation, first drafts, code generation, transcription, translations and templates,” Zoller continues. “Prohibited uses included confidential data, research and citations, data analysis and solution development. We instituted requirements for use, which involve review, verification, edit output, disclosure of AI use and brand alignment. In addition, we stated a commitment to equity in which we will not replace jobs with AI.”
“Most of the AI policies I have seen in firms revolve around risk awareness,” Leverett says. “Many concerns parallel those in the real world, like protecting confidential and proprietary information or recognizing intellectual property concerns. Also, AI is imperfect and errors may occur, so it is essential to confirm and verify outcomes whenever possible. I typically advise that firms consult theirattorneys and insurance carriers when making AI policy.”
“The first step should always be identifying a need,” King says. “What are you looking to achieve? What problems are you trying to solve? You may find that AI is the fix, or you may find that there is a better solution. Businesses can’t afford to embark on vanity AI projects just for the sake of it. There must be a clear need and a problem to solve. Once you know what you are trying to achieve and how you intend to use AI, you can build your policies around it.”
The Future of AI
We are just beginning to realize and utilize AI’s potential. The underlying technology is built for exponential growth in speed and computing power, so it is bound to evolve and change the industry and the world as it does.
“The integration of AI with other technologies in design, construction and building operation holds the potential to create a whole new level of interaction, information and performance,” Hollwich says. “These fields often operate in isolated silos but over time they will inevitably engage and merge. As we progress toward fabricating buildings through AI-enhanced processes, we will be able to quickly generate and evaluate designs we co-pilot with feedback as to their impact, feasibility and cost.”
“Right now, AI is limited by general domain knowledge,” Schroeder explains. “It has been trained on the internet, which is fast and has deep knowledge, but is not vertically integrated and doesn’t have domain expertise. The next generation of AI is going to be able to reason much more deeply. I won’t be just recalling information it has gathered. Next-generation AIs will learn the rules behind the data, which will give them new capabilities.”
No matter what the future has in store, AI is here today and poised to make a big impact in the short and long term. Whether firms are using it right now or not, it’s important to look ahead and develop effective strategies for proper use and implementation.
“One of the major considerations of AI adoption is the human cost,” King points out. “Introducing such a disruptive and transformative technology is going to be a major adjustment for all involved. You are likely to face resistance in various forms. The best thing organizations and their leadership can do is work with their people to bring them into the fold and make them active participants in this new chapter. If you want any momentum at all, you need to have buy-in. Make it clear what you plan to do, why you’re doing it and how it’s going to happen.”
“Approaching the use of AI with an ethical foundation is critical for responsible and sustainable business practices,” Hollwich says. “But we also have to remember to be playful and explore the creative and sometimes spontaneous use of this new tool.”