Builder Magazine: Artificial intelligence is one of the most talked-about topics in construction right now, and builders are feeling pressure to do something with it, often without clearly understanding how best to incorporate it.
At the same time, new tools launch weekly, promises are everywhere, and the pressure to adopt something to stay ahead of the competition is real.
7 comments:
When I took a couple Construction Management classes at my undergrad, we regularly talked about AI, and one of the more off-putting beliefs that one of the instructors voiced, and is echoed in this article is the belief in giving as much info as possible to the AI. For that instructor, he said that the single best step into preparing for our future construction careers was to adopt AI engines and start using it as much as possible so it has as much time as possible to adjust to your practices and you have time to fully integrate it. Something kinda similar is echoed in this article when they talk about making sure that you use a single AI engine across your entire business to prevent data artifacts, and provide consistency across the whole company. I think aside from the aspect of giving that quantity and depth of information that puts into a data center that almost certainly isn’t owned and solely maintained by that specific company, so there is theoretically a world where all of the company’s projects, and potentially client or employee information regardless of exclusivity, is held in the hands of large scale data centers. Aside from all of that, I think it also goes against one of the practices I remember reading about in one of the other articles a while ago about essentially having contingency plans. There should always be at least 3 avenues of completing a task. While integrating AI doesn’t inherently make those using it incapable of creating and/or managing the schedules, estimates, RFI’s, etc. it seems as though it could create an achilles heel of sorts within the company, if it were to ever fail.
This article is so true. The AI bubbles about the burst in that we all reckon what it’s actually good at what it is not very good at. This is one of those examples. I love that they brought in the line “garbage in, garbage out” to describe applying AI to an already broken process is only going to make it faster and more broken. The end of the article got right at the speed spot for theater for me. Putting aside for a moment the myriad of environmental concerns: the best use of AI is to save time on the administrative tasks, such that the people in our teams can do more of the tasks that should be done like humans. I want to say AI is not gonna take the job of artist for this reason. But suddenly that’s not the truth. Because people are lazy and cheap. Such that businesses have more than enough money to pay people to create their art, their concept art, their marketing, would rather rip off an artist to save a buck than pay a person for a service. AI won’t touch performing arts, but it is already wrecked havoc on visual arts.
- Mags Holcomb
The thing I liked in this article was its realistic approach to AI. The phrase that caught my attention was the term "multiplier, not a foundation," used by Copenhaver to describe AI. This concept goes far beyond construction as well. When your logs are not done properly and schedules are estimated, AI will help you do them even faster but won't improve anything else. Also, it was quite interesting to see the mention of "wrapper" products with the nice interface and model that anyone can use. I think it is worth asking each vendor what the product is really theirs. Of course, I have been reading the article with some degree of skepticism because of the sponsorship from Buildertrend and huge numbers like 97% of time savings provided by the company. However, all the suggestions in the article were true: find the problem, improve the process, and then automate it. The part that I liked the most was the end when Copenhaver mentioned that he wanted AI to get rid of all busywork for his team to spend more time in the field with customers.
AI will continue to be an incredibly tricky topic but I think this article does a good job at finding ways that it is actually useful for a job that creates purely physical outputs. Using AI solely as a data management operation allows it to be an incredibly personalized software that any company can implement. The AI isn't really doing much thinking or theorizing new ways we can implement construction techniques; it's just taking basic information that is already digital and sorting it into a readable and referenceable database. Personally I find it challenging to implement AI into my own day to day life because I feel as though I should be learning and practicing, I can see the benefits of using it in larger businesses as a database. Although giving an AI engine all of your company's information and just letting it have free range without restraints is slightly terrifying.
I am glad that AI hasn’t been so rapidly accepted into all industries without consideration for how it can affect their workflow. For an industry that is so old but varied, I think AI will have a hard time integrating. Without a solid understanding of construction and all the complexities, AI will just lead to more work for the companies. This can carry over into theater because of our similarities in problems to be aware of and complex but individual systems where all the information is rarely placed in the same location. I believe that it is because of this that AI has such a hard time integrating into real-life and physical industries where the system has not already been semi-automated and organized onto the internet. I think when AI is forced into spaces like these, it can perpetuate small-scale issues further into the system, its people, and its outputs.
I feel as if we’ve been hearing a lot in recent years that utilizing ai is a skill you need to have to in order to be a competent employee in the modern era and while that does have some truth to it I feel like we’re kinda of overestimating how useful ai is at this current point in time. Yes ai has the potential to make your workflow faster by assisting you with a lot, I think at this point in time it’s still kind of dumb and actually doesn’t know as much as people believe meaning it doesn’t have the ability to do a lot of things. Yes while ai has access to an unbelievable amount of data and information it doesn’t actually understand it that’s why it has the potential to make mistakes or give you wrong information which is why I don’t think it should be a main tool in a field where it’s important to be precise and safe like construction
I think there is an interesting point that was made in this article. It’s the fact that AI can’t make a process better or fix it, it will just streamline it as much as possible. It also doesn’t have the full scope and context to a business or a jobsite that the people do. In order for it to be a useful tool you would have to make sure that it had all the information and was made to fit into a construction context. Like theatre there is so much knowledge and tricks that people have picked up that a regular AI model might not understand. Having one that is tailored to your working environment would fix that problem but it also means a lot of preparation and specific data that has to be given to the machine in order for it to actually understand what’s going on and be able to help more. Overall, I think it’s interesting that there is AI already be used in physical work places like jobsites, but I also appreciate that it can’t really take over there yet because it such a person forward job.
Post a Comment