There is a concept in engineering called drift. This is when systems or environments slowly change over time away from their original purpose and configuration.
AI risks doing the same to brands, perhaps in a much more treacherous way.
Configuration drift is detectable. The downstream impact of a brand losing its voice may never be.
Brands are defined by some quantifiable and qualitative set of metrics that may be difficult to assess individually, but collectively, create something unique. They span the fonts that are used, the stories that are told, the positioning of the brand, and many other factors.
These are often forward-looking, a result of the human spirit and creativity combined with market research and understanding a target market.
Unrestrained AI use does not just work outside those definitions; it begins to create its own: what the brand should sound like, what the campaign should be, what the secondary color is. Systems are what keep AI operating in a bounded way, while still moving faster.
The questions worth asking are:
- In what part of the process can you get the most out of AI: the greatest upside with the least downside?
- How close to the finished product is AI allowed to get? Think in terms of writing, visuals, and code, though each one is weighted differently.
- Is training AI on a brand voice enough to let it run unsupervised, or is a voice distinctly human?
A brand’s voice, look, positioning, and more are at risk when unrestrained AI is used to speed up the process. Just a little bit here, a little bit there, and over time, the brand is brought more into the middling average.