Kaplanoglu
The construction industry faces challenges due to rapid growth, labor shortages, a retiring workforce, declining productivity, increased complexity and safety performance challenges. Many people are looking for solutions to these problems using artificial intelligence (AI).
You can say we’ve seen it before: the workforce will adapt, and AI is a hype anyway. This time, it’s different, and you have to look beyond the hype.
Speed of technology advancement
Let’s take a look at how fast technology is improving. In the design services industry, the use of image generators is gaining traction for early stage design. It takes weeks for a new and improved image generation model to become available. Three weeks after Google’s revolutionary Banana Nano model, ByteDance Seedream 4.0 released a superior model. AI models are context aware; you can request a change to an image and see it updated in real time within minutes. The vast amount of research, funding and advancements is accelerating the pace of progress.
Availability and adoption at scale
We are experiencing change differently because anyone can access AI online. In the past, it took decades to implement technologies. This time the adoption of AI is much faster than any other technology due to its wide availability. ChatGPT gets more queries per day than Google Search got in its first decade. OpenAI reported that it had more than 700 million weekly active users and more than 3 billion messages sent daily as of August 2025.
Efficiency gains
It has not been easy to measure the impact on the construction workforce. Last summer, we started hearing that companies weren’t replacing employees who were leaving or retiring. We now have proof that by 2025 there was 30% less hiring of university graduates across all sectors. While the overall unemployment rate was 4.3%, the rate of recent college graduates was 6.5%, the highest in more than a decade. Some industries are already changing rapidly. Data shows that more than a quarter of all computer programming jobs have disappeared in the past two years, and employment in this business is expected to decline by 10% over the next decade.
Naysayers will point to an MIT study that found 95% of AI implementations fail. The study focused solely on generative AI, the kind that necessarily creates new code or other content, and had a relatively small sample size. I agree with one of the findings of the study, which showed that change management was a key issue. The results align with my own observations: successful AI deployments either use AI-embedded products or don’t go it alone and partner with others. In my opinion, buy AND partner are the two correct options.
Some studies also show that people often don’t say they use AI because they think it devalues their work, even when they actually use it. On a personal level, people report significant productivity gains, but we struggle to measure the impact of generative AI on business outcomes. I think generative AI is useful in isolation, but it really becomes powerful when working with AI agents designed to perform specific workflow tasks. We will see more studies measuring the positive impact of use cases that leverage generative AI and AI agents working together.
AI encompasses a range of technologies, including machine learning, deep learning, computer vision, general AI, multimodal AI, and AI agents. For public education, the capabilities and impact of generative AI generalize to AI adoption as it is the most visible and accessible.
Like any technology, generative AI can be successfully implemented when managed well. IKEA implemented a call center bot without laying off employees. They retrained 70% of them to become design consultants and the rest continue to support the call center. By retraining call center agents, they made billions in additional revenue selling their products while offering design services. The bot use case can be applied to our industry by providing insight and guidance for the day-to-day operations of our workforce. It’s also helpful to have employees who stay in customer service call centers for all the issues that a call center bot can’t just solve.
I’m sure the efficiency gains will become more measurable as technology and adoption continue to evolve, and we reskill our workforce.
Now that you have some perspective on the current state of AI, let’s take a look at the potential impact of AI on the construction business over the next decade.
Jobs (not Steve)
I think construction jobs will become more attractive as jobs in other industries become more automated. Due to the physical nature of the work and its complexity, I predict that the automation of our industry will be delayed compared to other industries and that more job seekers will move into our industry.
Workload management
AI infrastructure companies announce investments totaling hundreds of billions of dollars weekly. We can’t keep up with the ads. Some believe we are experiencing a bubble and work will decline in 3-5 years. I think the opposite: the work will continue to grow as the AI infrastructure we’re building today will need to be updated as technology advances. We will also build new infrastructure, such as roads, schools and hospitals, to enable the gains from technological advances.
Stay able
We will also see changes in the way work is done. The ability to do more with fewer people for engineering, design and construction will help smaller companies compete with larger companies. Companies will have fewer people with specialized skills.
Construction labor is an example. There has been a significant improvement in robotics in the last 12 months. Construction has been listed at the bottom of the AI impact list due to its physical nature and complexity. Still, I see a clearer path to higher levels of robotics adoption in industry. For construction jobs that require physical skills, advances in robotics will help augment (not replace) the workforce and fill gaps. The adoption of automation in large and remote projects is already driving productivity. This change is likely to happen within a decade in today’s world.
Making sense of the business
Finally, construction business models will change. We’ll start measuring value, instead of running to the bottom. We will charge based on the final value of the services, not by the hour.
I believe these changes will occur in the next decade. The good news is that other industries will transform quickly, giving the construction industry more time to react and adapt. We will also have two different speeds. Because of digital workflows, architecture and engineering will transform faster than site construction, which requires physical workflows.
We have to keep our eyes on the ball.
Burcin Kaplanoglu is a construction technologist and AI expert. He is active in industry organizations and on LinkedIn where he provides educational content related to technology, innovation, robotics, AI and industry use cases.
