ChatGPT is not part of an AEC AI Strategy. Your AI Strategy is likely non-existent or wrong.

Last week I gave a keynote address at the PSMJ AEC Innovate Conference in Boston. My “This Changes Everything” keynote focuses on the skills that we as an industry need to develop to thrive in an age of AI. My book is based on this talk that I have been giving to awhile now. PSMJ did a great job of having sponsor introduce speakers vs sponsors giving keynotes (pay to play). 

My key take away is that the AEC industry is way too focused on AI equating to LLMs. The reality is that while LLMs have a place in the AEC industry, I do not believe that they are the front a center of applications that will disrupt the AEC industry. I really believe Computer Vision, Machine Learning, Expert Networks and Autonomous Robots will be the disruptors. Why? Because we deal with visual and physical things in our industry. The AI tools that will accelerate “Idea to Execution” will be less about LLMs. 

If a Civil Engineer has an “idea” about a bridge design, it will be about taking the idea and putting it into a visual format (drawings/3D) to communicate it to others on the design team. Once iterated, it will be about pushing the design to a contractor to physically build it (robots and people). These designs are translated into instructions by the contractor (means and methods). These instructions may be given to a person or may be translated into motion control for a robot. 

Why is this disruptive? The “iteration”. There are a lot of hours to develop the design and then to convert it into instructions. 

Concept-Schematic-Design Development-Construction Documents (Drawings and Specs) then translated by contractors into Shop Drawings, Submittals, Work Breakdown Structure, Estimates (Quantities and Labor)..from the GC to the Subcontractors to the Trades to the Labor (in most cases, People but also Robots). 

So where is the disruption? Let’s not call it disruption for a moment and call it Value Capture. 

Effective value capture is a key driver of profitability and long-term business success. Companies that are able to create significant value for customers and then retain a large portion of that value tend to be the most profitable and sustainable. Managing the various levers of value capture is a central strategic challenge for businesses.

In an AI driven future, what if an Idea is automatically iterated to the customer’s needs, budgets and schedules created, and then pushed to labor/robots to execute? What if this is weeks and not months? Are there some steps I’m missing? Probably, and when will this be feasible? Maybe not tomorrow, but soon. 

Then begs the question, WHO will be the entity that captures this Value? Any of the current players, or maybe a totally new player. 

I have always said that our industry looks more like the movie industry than manufacturing. A bunch of people show up, work on a project. A “producer” beats on everyone to get things done for cheap and then everyone leaves. Who is the producer in the AEC space? Unfortunately, sometimes the design team and sometimes the General Contractor. Rarely the owner. 

I was on a flight and a Hollywood stuntman was sitting next to me. I asked him this question. He basically said, this is what the strike was mostly about. That AI and streaming were killing the business and that the producers were capturing all of the value. That his role is always at risk from being “CGI”ed. But CGI used to be pretty expensive at times versus stunt men. He says, that’s no longer the case. So what is he doing now? Trying to write his own screenplay using AI, creating story boards using AI, has become versed at Sora and hopes to produce his own movie. IE, he is moving upstream to capture all the value. 

So while you are tinkering with try to use LLMs to respond to RFPs, write emails, create social media posts, job descriptions and the like. (These are all great tactical use cases). How are you building an AI strategy like my new stuntman friend that is a Value Capture strategy?