For all the seemingly magical things artificial intelligence can do, there is one thing that still baffles it: generating images of wristwatches set to 6 p.m.
At first, it seems perplexing that technology capable of generating full-length, authentic-sounding novels in minutes could be stumped by this simple request. However, it makes perfect sense when examining what AI actually does.
“Every advertising image out there of wristwatches shows the hands at 10:10, because that’s the most attractive way to display watches in advertising,” says Dr. Brian Spaid, marketing department chair. “If there are no images out there of what 6 p.m. looks like, then the AI is going to give you pictures of what it looks like at 10:10 and think it’s complying.”
This quirk in AI is an example of the technology’s human-driven biases, something that marketers must be aware of before it manifests in larger ways. While Spaid integrates AI into his classes, he is also mindful of its many limitations, and how marketing students must adjust to account for them.
DOES THE MACHINE KNOW US?
An adept marketer is grounded in fundamental understanding of people: what their needs are, what motivates them and how a product can best fit their lifestyle. These skills are honed through interviews, focus groups and other experiences that put the marketer in direct contact with their customers.
Artificial intelligence can be a powerful tool for identifying trends and patterns in this data set. However, Spaid warns against using AI as a shortcut around interacting with people.
“The marketer that relies on AI to give original consumer insights is being very lazy,” Spaid says. “Artificial intelligence is only going to tell you what it has access to in its database, so there’s always going to be a time lag in measuring consumer sentiment. If you ask an AI about reaction to the California wildfires, it won’t know because it won’t have caught up to the news cycle.”
Spaid advocates a blended approach, using time-tested marketing research to generate good data, then applying AI tools to interpret the data in useful ways.
BUILDING SKILLS OFFLINE
Spaid and his fellow professors face a dilemma with the advent of widely available AI. On the one hand, it seems illogical to ban a tool in class that employers will not only allow them to use, but encourage them to use. On the other, students must still gain conventional marketing skills to be effective professionals, which over-reliance on AI can inhibit.
“You can ask AI how to do heart surgery and that won’t make you a heart surgeon,” Spaid says. “You may understand the words that it’s telling you, but if you don’t understand the underlying concepts of, say, segmentation, targeting and positioning, using AI is not going to help you as a marketer. Students still need to learn the information to use the models appropriately, and that’s what we as professors can give them”
Ultimately, Spaid envisions broader adoption of AI; its ability to generate profits and increase efficiency is too great to ignore. He does, however, caution over-reliance on these tools.
“Students need to at least know enough to know if the AI output makes sense.”
LOOKING TO THE PAST TO AID IN THE FUTURE
When contemplating how to deal with artificial intelligence in the classroom, Spaid sees parallels in how academia reacted to the rise of the Internet and Wikipedia in the 1990s.
“All of a sudden, the floodgates opened on free information and higher ed was really caught off guard about how suddenly it happened, and I see similar reactions happening now,” Spaid says, while also recalling that in his previous career as a business-to-business web and application designer, clients tended to expect a higher level of work whenever a new technology became broadly available.
However, Spaid does see a crucial difference between today’s breakthroughs and those of the past: AI can complete work on people’s behalf, whereas the previous era of technology only catalyzed access to information.
He also believes that eventually, artificial intelligence technology will mature, and its rate of growth will slow, leaving some kind of equilibrium between humans and machine in its place.
“There’s a very short window for firms to unilaterally take advantage of technological advances like these until the balance shifts, as it always does.” Spaid says.
