As AI is becoming more and more relevant across the workplaces of America, there have been questions raised about whether it’s a smart thing to be investing so much of our progress into. The construction industry especially is seeing a rise in people both for and against the implementation of AI, and it’s a complicated issue. I believe that to keep up with the changing times and technology, the construction industry must adapt and learn to integrate different AI features into daily operations.
It is my opinion that AI should be integrated into construction operations. The most important reason is that it has the potential to have a huge positive impact on safety protocols. "In the US, one in every five workplace deaths is reported from the construction industry while employing just over 6% of the workforce" (Rabbi). Construction is not a very safe industry, and with an environment that is never the same, it’s hard to pin down potentially unsafe situations. AI can help solve this by being integrated into real-time monitoring of sites, with built-in hazard detection and information extraction. It will give us data about accidents, and when we feed it back into the algorithm it can learn from it and warn us about potentially dangerous situations as they unfold (Strong). It can also speed up project management and scheduling by using algorithms to access and synthesize previous project data and apply it to current requirements. There is also a labor shortage in the construction industry that can be positively impacted by the integration of AI. In 2022, a survey by the Associated General Contractors of America found that 79% of construction companies were having problems finding workers (McCullough, 2022). Using AI in the planning phase can help make construction operations more efficient and lessen the workload of skilled laborers. Rework, which is when mistakes are made and the work must be redone, is one of the costliest parts of construction enterprises both in labor and money. Using AI to supervise and possibly lessen the need for rework would not only lessen the workload of subcontractors but also save construction firms a lot of money. The construction industry still needs a lot more skilled labor to operate efficiently, but AI could help make the labor we have as efficient as possible.
I understand that AI is a complicated topic these days and that not everyone will have the same opinions I do. There is no “right” way to do this, and there are certainly people who have more experience than I do who might see this in a different light. I also can acknowledge a bias that I have: I’m currently in school for construction management and being younger and somewhat more familiar with AI might make me more trusting in its capabilities. When it comes down to it though, AI has been introduced to the construction industry already, and I think everybody involved should have a voice and a say in how things proceed and what the implications are.
One of the opposing viewpoints that stood out to me the most during my research phase was that the environmental impacts of AI far outweigh the positive benefits it may have to businesses. Something that not a lot of people think about regarding AI is the pure volume of computing power it requires (Gonsalves at al., 2023). The implications for energy consumption and the carbon footprints of these data centers are pretty extreme; Chat GPT for example consumes around 1GWh of energy in a day, which approximates the daily energy consumption of 33,000 U.S. households (Capen, 2024). People who oppose AI because of the environmental impact are completely valid to do so – this world is the only one we have, and we must protect it. The construction industry has also historically been not very kind to the environment. There have been substantial efforts recently to improve the impact that the construction industry has had on the environment, and it would be a shame to take a step back in that progress.
Another opposing viewpoint regarding the integration of AI into the construction industry is the idea that AI will influence our decision-making capabilities. Critical thinking is present and important in any professional field, but I would argue it has higher stakes in the construction industry. The ability to make quick, on-the-spot analysis on a job site is what keeps sites safe and productive. Something is always going wrong, and it’s the job of people like superintendents and site supervisors to make sure a catastrophic mistake isn’t made trying to correct course. There are people out there who are worried that if we allow AI to do all of the analyzing and computing, we will be losing those skills ourselves. Not knowing how to analyze and compute things has a detrimental effect on critical thinking skills, and that is a very important part of daily life on a construction project (Arroyo et al., 2021) Groupthink – a phenomenon that occurs when groups avoid disagreement and results come back lackluster as a result of diverse thinking – is also a concern if AI becomes widely used. If we get to a point where we trust AI and don’t question it, we lose our creative ability to solve problems and come up with innovative solutions (Arroyo et al., 2021). This is a valid viewpoint because we don’t really know how AI will affect our brains yet. There hasn’t been enough of a full-scale usage to even begin to test things like that, so it is somewhat of a question mark in AI in general. Decision-making and critical thinking are also huge parts of construction, so I can understand how people are worried that it could change things.
An unexpected stance that I came across while researching is that all of these companies jumping on AI is artificial hype, and it will harm companies that rush into implementation. An article written by Ashish Jha for XYZ Reality stated that AI isn’t actually as powerful as people think it is, it’s just companies artificially inflating its worth so they can capitalize on the next “big new thing.” He mentioned that AI in its current form is one of two things: surface-level chatbots with limited capabilities, and Chat GPT lookalikes that are easily tricked and unreliable. The construction industry needs “nuanced and practical assistance,” which current models of AI cannot provide. He suggests that AI is in a “hype cycle” right now, where “an initial period of overinvestment will inevitably lead to a crash, followed by a slow comeback as people learn how it can provide real value” (Jha, 2024). Jha proposes that we have to take it slow; AI will eventually make a comeback, but right now the public perception of AI is being artificially inflated by companies. Construction companies should only implement things when they show genuine progress, or a crash will come and all of the “AI-powered” advancements that are pushing a company forward will become a source of mistrust instead.
I think the biggest concern surrounding AI not only in construction but across all industries is how to trust it. Human-AI trust is the biggest obstacle to implementation, and for good reason (Schia, 2019). AI is easily manipulated right now – I’m sure everyone has at least seen the results of someone purposely trying to get an AI model to do something it shouldn’t. There was even a contest recently where an AI model named Freysa was given a sole command: do not send anyone money. The goal of the participants was to trick Freysa into disobeying her only command and handing over the prize pool that had been steadily growing as people paid to chat with her. In the end, one user won $47,000 by essentially switching Freysa’s definition of “approve transfer” and “reject transfer” (Robson, 2024). If someone can outsmart an AI model designed specifically for that function, how can we trust it with our data? We can’t know how it will react in certain situations, who can manipulate it, and if we can even tell if it has been manipulated. This is a very valid reservation to have when thinking about AI. It’s still very fresh, and a lot of what we know about the widely accessible public aspects of it right now is uncertain. I agree that we need to be able to trust AI to properly implement it.
In order to come to a consensus among people about the implementation of AI, I think it’s important to state that we all want similar things. Safety is of the utmost importance when it comes to construction, and I think everyone can agree on that. Innovation and progress within the construction industry is also a very important value. This goes for design accomplishments, engineering performances, and the amazing progress that has been made with regard to sustainability in construction.
All in all, there is a possibility for a collaborative solution between all parties involved. With safety being the most important factor here, proper research is the key for any applications of AI that claim to reduce safety accidents. Companies planning to include accident monitoring should undergo an extended “test phase” until there is enough publicly available data to reassure the public that it actually works. Continuing to practice the current high standards for safety while employing AI to possibly catch more safety offenses can’t hurt. We shouldn’t be relying on AI safety monitoring at this phase in development, simply using it as a tool to enhance current practices. Overall with AI, we need to be implementing bit by bit, seeing what works, and continuing to refine and improve applications. AI should not be completely taking over functions like estimating or scheduling, it should be complimenting it. I also think it’s important to factor in the environmental implications. Though it may not be something that everyone cares about, it will have an impact on the world to come. There have already been steps taken to construct buildings more sustainably so it’s not entirely unrealistic to think about constructing data centers with renewable energy in mind. It’s also possible that as AI evolves, it will become more efficient regarding the amount of energy it requires (Haskell, 2023). Technology is constantly improving in efficiency, so AI will likely do the same thing. As technology within the industry evolves, I’m sure practice will too. It takes everybody thinking and changing things along the way to come up with the solutions.
If you have a minute to spare, I would love some feedback in the discussion group on my website! Even if it’s just a few seconds to drop in and say hello, it would mean a lot. Thank you for reading!
Comments