Thomas Davenport and his colleagues have a recent piece in the Journal of the Academy of Marketing Science on Artificial Intelligence, AI, and Marketing’s Future. This is an eminently readable piece, which I appreciate. It lays out some issues with AI and where the marketing discipline may go.
AI And Human Beings
I am fascinated with people’s responses to AI, see more here. I think it is interesting that people expect more of AI than they expect of people. When a person goes wrong we might be more more forgiving. Take the common example of AI self-driving cars.
“…customers impose higher standards and seek zero accidents from AI. Understanding the roots of this excess caution is important”Davenport, Guha, Grewal and Bressgott (2020) page 28
I totally agree that this understanding is critical. Should we write this off to consumers not being very bright? (After all human caused deaths in road traffic accidents are just as bad as AI caused deaths). Or is there some level of reasonable caution about new tech that we should adopt? What then is the right level of reasonable caution?
This has crossover with discussions of GMO, Genetically Modified Organisms. Advocates are surely correct that the demonstrated problems are low and the benefits are large. Yet, it is hard to show that unknown risks are not massive because, well, they are not known. You can’t really prove that there won’t be massive future disasters. (Sadly I’m not giving you an answer to this).
Helping Consumers: AI And Marketing’s Future
How then will AI impact marketing’s future when it comes to helping consumers? I am pretty optimistic to be honest. There will undoubtedly be challenges. The authors note:
“…if consumers’ ideal preferences actually differ from their past behaviors (e.g., consumers trying to stop eating unhealthy foods), AI might make it harder for them to find and move towards their preferred options , by only presenting them with choices reflecting their past behaviors”Davenport, Guha, Grewal and Bressgott (2020) page 37
This is a worry but marketers already steer consumers to unhealthy choices without the need for AI. For example, marketers do this through SKUs stocked on the shelves in response to non-AI assisted ideas of ‘what sells’. At least consumers may have the ability to share their diverse, and somewhat conflicting, preferences with increasingly sophisticated AI.