AI and competitive advantage can work against each other when companies mandate AI use without creative guardrails. Two new studies show that the same generative AI tools boosting individual output also homogenize the ideas teams produce. Over time, that erodes the differences in thinking that give firms their edge.
Years ago, an executive at a Fortune 500 company told me, “If I could use AI to amplify my A-players and get my B-players to operate as A-players, I'd buy that product in a heartbeat.” Many companies are now treating AI models from Anthropic, OpenAI and Microsoft that way. The pitch is speed, consistency, standardized processes and higher output. But mandates without creative guidelines may reduce the variety of new ideas inside teams.
What Studies Reveal About AI And Competitive Advantage
Two separate studies, Doshi, Hauser, and Meincke, Nave & Terwiesch, have shown that AI levels up the overall output of teams within a company in terms of quality; it simultaneously homogenizes the ideas in those teams. Part of this is due to generative AI being good at incrementing on existing ideas rather than creating net new ideas according to the study “ChatGPT decreases idea diversity in brainstorming” by Lennart Meincke, Gideon Nave & Christian Terwiesch.
If a critical mass of companies in a specific market begin to use the same set of models, it could chip away at whatever competitive advantage they had in their market. This decreases the distance between firms' products, services, and performance.
What NVIDIA Teaches About AI And Competitive Advantage
In his book, Competitive Strategy, Michael Porter said that competitive advantage comes from differentiation. An example of this can be seen with NVIDIA. Many years ago, NVIDIA decided to focus on GPUs when the rest of the chip industry was focused on CPUs. Focusing on CPUs would have put them in the most crowded part of the market where Intel was already the dominant player. NVIDIA saw gaming as a major emerging use case that was going to rely on parallel processing in chips. This allowed them to focus on GPUs when no one else was. Because of these decisions on strategy decades ago, they are reaping the rewards today.
Now, imagine a software program that NVIDIA had years ago that helped them automate work and come up with new ideas. Let's imagine that the majority of their competitors used the same or similar software. If that software ultimately caused NVIDIA, Intel, and other major chip competitors to have similar ideas, it's possible that NVIDIA would have picked a very different strategy and not pursued GPUs. Instead, they may have focused on the path of winning with CPUs. They still may have been a major player, but probably neck-and-neck with its competitors, and possibly would have missed out from the gains of having a lot of experience producing GPUs by the time generative AI accelerated chip demand.
How To Use AI Without Losing Your Competitive Advantage
It comes down to how you use these tools so it doesn't over-homogenize ideas to a point it steals your company's ability to differentiate in its strategy. There is a growing trend in companies today to mandate AI usage, with the only success metric being based on usage, like Meta's now-defunct token leaderboard. This has the potential to erode competitive advantages within industries if token spend remains the north star. Companies should be taking an outcome-based approach. It's especially important that employees can explain their thinking and still leverage their own creativity even when working with AI.
Companies will see real productivity gains from AI. But those that treat usage itself as the goal risk a quieter cost: their teams start converging on the same answers their competitors are getting.
The firms that hold their edge will pair AI adoption with explicit room for human judgment and original thinking.