Factories across New England swapped steam engines for electric motors in the 1890s, waiting for productivity gains that took thirty years to arrive. The breakthrough only happened when factories were torn down and rebuilt with assembly lines and individual motors at each station. A recent piece from venture capital firm a16z asked where the value went after AI made every individual ten times more productive without making companies ten times more valuable.
Swapping the motor, not the factory
The answer, according to the report, is simple: we swapped the motor but did not redesign the factory. The authors frame the issue as a failure to rethink the underlying process rather than just applying new tools to existing workflows. This concept is the central theme of Episode 4 of The AI Edit, where Samantha McLean applies it to a familiar real estate agent process: the headshot.
The episode breaks down a standard workflow for professional photography and shows how changing the starting question alters the result. The lesson is that doing the same thing faster is valuable, but it is not transformational. The real gains come from rethinking the process entirely before writing a single line of code or using a single tool.
Related: Grace Anderson and David Sylvester open Ray White Katikati
McLean introduces the “AI First Formula,” which starts with one question: “What could AI do here?” This question must be asked before the traditional process begins. When the constraints of time and cost are removed, creativity follows. The ability to iterate endlessly changes the relationship with the output in ways that speed alone cannot achieve.
The logic extends beyond headshots to every process in a real estate business. If the starting point is a traditional method, the result will be a faster version of that method. If the starting point is the question of what AI can do, the result is often a new process entirely. McLean built a free AI headshot generator to demonstrate this, offering 12 styles across Professional, On Site, and Fun categories.
Rethinking the workflow
Applying this formula requires patience. It is easy to see a new tool and immediately bolt it onto an existing workflow. The friction of learning the new tool often obscures the opportunity to redesign the workflow. The episode suggests that the most productive use of time is not to master the tool, but to ask what the work should look like without the constraints of the current method.
Related: Gavin Rubinstein reaches elite status early
Creating a new workflow often feels messy at first. The standard method is comfortable because it is familiar. The AI method requires a willingness to let go of that comfort. It involves testing ideas, failing fast, and adjusting the process based on what actually works rather than what is expected to work. This iterative approach creates a feedback loop that speeds up development and improves the final product.
Looking at the history of electrification, the factory redesign was not a minor adjustment. It required tearing down the old structure and building a new one. The same level of commitment is required for AI integration. Without that structural change, the gains from AI remain marginal. The tools are powerful, but they are only as useful as the systems they operate within.
