Specificity for marketing vs. replaceable by AI.

Imagine a gel that is sold in the market that says, “It will shine your shoes, make a great sunscreen and also works as a great flame retardant!”

Would you buy that?

I don’t think you would. It’s too weird of a mix of things. Even if they are all true and you were the creator of that product, you would do much better creating three different containers with three different labels for shoe polish, sunscreen and flame retardant even though they would all be filled up with the same substance. Marketing seeks specificity, and not only does it seek it, but by finding the perfect niche people have made fortunes.

However, the world has a new force approaching, which is artificial intelligence (AI). In Artificial Intelligence (AI) training data is used in order to develop an unstructured model, one where the programmer doesn’t have to provide an equation in order to classify or predict something. This can range from recognizing images, to predicting stock movements, and more. In the future, it’s likely that if you build the appropriate AI model and provide it with the proper training data, it could even create art of a certain style after being trained by a data set. I wrote about that previously here.

There is going to be tension here between these two concepts. Good marketing demands specificity. AI seeks to replace people doing specific work. How can someone find the gaps that will remain between them?

Here are some thoughts at the current moment, but changes in technology could negate this:

  • If you work in a narrow field but combine a broad range of skills, you are probably safe. Interior design for example combines knowing product availability, spatial cognition, color choices, client preferences and more. It’s a narrow field, but diverse in skills.
  • If you work in a broad field, it’s possible that AI will empower you to split it into niche markets each with their own marketing by allowing your marketing efforts to scale by using article and image generators for example, which previously would have required you to manually figure out.

There is so much going on in my mind about this at the moment, it’s hard to communicate it all. I’ll write more about it in the future.