The Challenges and Opportunities of using GenAI: Business and Education

Generative Artificial Intelligence (GenAI) is a very impressive general purpose technology: it can write poetry, recipes, cover letters for jobs and (after several suggestions were rejected) these images. The people who developed the foundations were recently recognized with a Nobel Prize.

Every new and old business seems to claim that it is using AI: e.g., Coca-Cola has offered a drink with “Futuristic Flavor and AI-Powered Experience”. Despite the rosy predictions, people are starting to ask where is the revenue produced by this investment? So, we seem to be entering the Trough of Disillusionment phase of the Gartner Hype Cycle.

GenAI is also forcing instructors to rethink what to teach when it can score highly on standardized tests in economics. Even so, business students seeking to profit from this technology may be at the “Peak of Inflated Expectations” if they think profit is an outcome rather than recognizing that profit is an incentive.

First, as context, reports say that world-wide spending on AI is over US$200b (billion) per year now. That number is puny compared to the US$5,000b spent on information technology in general.

As a general purpose technology, the value added of GenAI depends on complementary investments. This feature of market forces is very old. Electricity is also a general purpose technology and nobody thinks about it (until there is a power failure). People value the things which use electricity, such as toasters, TVs and the computers used by the AI providers.

Generating revenue from GenAI depends on matching the technology with complementary investments appropriately. As I read more, I notice that the media emphasize what “AI can do” (i.e., amazing) but companies will succeed by balancing costs and benefits on each margin (e.g., features, human support, …). The profit-maximizing rule “marginal cost= marginal revenue” is a reminder that the appropriate point of balance will differ for different companies.

It will take some time to find what benefits consumers the most. Embedding a camera in a phone did not cause the bankruptcy of Kodak immediately [1] [2]. Now, the benefit of having a camera in a phone is evident to everybody, in the convenience of taking a good picture anywhere anytime plus the ability to send the picture to others electronically, plus the ability to add funny masks or better makeup. Gains from trade exist because the value of a photograph is not the image or the chemistry of film-processing or solely in the lower cost and price of photos.

Gains from trade indicates that an innovative company might earn profit, and not just by charging a lower price than competitors. Products and services which embed AI will differ in unexpected ways: consider the puzzlement of humans from the 23rd century (Star Trek) trying to use an “advanced” computer from the 1980s.

I suspect that education will adapt in the same ways. Teaching students when the War of 1812 started or what is the sum of 43 plus 27 seems pointless when technologies can find answers faster and more precisely than memory. Maybe, teachers can skip the details and talk about more important and complicated problems. I suspect not. Calculators have been used in classes for many decades but students still make typing errors and forget whether 43 and 27 should be added or multiplied: technology enables students to find a better answer, but only if they ask a good question.

This perspective may be especially memorable in an economics class where, according to many people, the answer is often “It depends on …” Knowing what an answer should depend on helps to use GenAI effectively: it enables people to question silly answers and to find better answers. Successful students will match the new technology with this skill.

Now, it is your turn to write.  

• In 10 years, what will be the most important GenAI innovation embedded in an everyday items or activities, such as found in a kitchen or a hobby? What complementary investments will be needed to make it successful?
• Which types of jobs would be affected least by GenAI? Is it the job or the job skills which will change most?