The AI Disruption
Some will win, some will lose, the rest of us will sing the blues
Artificial Intelligence. It’s not so much coming as already here. Software programs like ChatGPT can ace college-level exams and write decent articles. So far these programs are mostly a novelty. They can be used to help you write a wedding speech or thank you note, but aren’t providing much help on a daily basis. That is going to change. In the same way that text messaging took years to take off after becoming widely available (at one point it was considered mildly impolite to have a conversation over text - why not just call the person), AI will likely transform from being a curiosity to a necessity over the next 20 years.
And with any transformation, there are going to be winners and losers. Some industries seem poised to undergo major upheaval. Many jobs will become redundant. Other industries be able to leverage AI and become much more profitable. Some commentators are speculating that AI could have a larger impact on the economy than the internet. So it was not great to come across the below graph this week:
The two-by-two grid is simple. The jobs in the top half are those that will have a high degree of AI impact. Those jobs in the bottom half will have a low AI impact. “Low/High Labor Shortage” is odd phrasing, but the jobs on the left are those that currently have enough workers to fill available jobs, while the jobs on the right do not have enough workers. According to this chart, the jobs in the bottom right corner are the most insulated from AI - there is a labor shortage and the impact of AI will be low. The jobs in the upper left-hand corner are in trouble - there are plenty of workers who want to do those jobs and they will have a high AI impact.
And I’m an economics professor.
Thank God I have tenure.
But in all seriousness, I don’t have much faith in a lot of these AI predictions. There are some that seem pretty solid. I think some jobs, such as paralegal, are going to have massive upheaval. A lot of that occupation consists of sorting through documents. This is the exact thing that would be a breeze for an AI. But that isn’t to say that paralegals are doomed as a profession. Often technology changes jobs rather than obliterates them. Some jobs, sure. Tractors did destroy the vast majority of farming jobs in the United States. Of course, tractors also made it possible for 1 percent of individuals to grow enough food for the other 99 percent and enable the rest of us to do different things. Computers did doom the typewriter repair industry. Will paralegals cease to exist? Doubtful. The best advice is not that paralegals should find a new job, but that they should leverage new technology to make them more effective at their current one.
Also, a lot of career upheaval predictions often turn out to be wrong. In the 1990s, many (accurately) predicted that the internet would change the American economy. But a lot of predictions of how it would change the economy were laughably wrong. There was speculation at the time that the internet would kill, or at least remake, American cities. This made sense. US cities had been declining for the second half of the 20th century. Then a new technology comes around that enables cheap and efficient communication of messages and documents across any amount of space. The natural prediction is that cities are finished. Of course, the opposite happened. Instead of a continued decline, cities rebounded in lock-step with the rise of the internet. Whether or not this was a causal change rather than a correlation is beside the point; the internet did not cause the end of the American city as we know it.
Are economics professors doomed because of AI? I think not. People have been predicting the end of professors for decades, likely because it is such a hard job to get. Contrary to popular belief, however, becoming a professor has never been easy. When I was an undergraduate, I was talking with one of my professors about the job. He surprised me by commenting on how tight the job market for professors has always been. Even when he got his PhD in the 1970s, it was a very difficult field to break into. If Covid taught us anything, it’s that the best education occurs face-to-face. All the technology in the world can’t replicate having an expert teach students in the same room. The role of tenured professor may be declining, but it isn’t because of AI.
Additionally, AI isn’t as good (yet) at tasks one would think would be right in the AI wheelhouse. For example, a friend of mine opened up a boutique a few years ago. It had a lot of unique gifts that were mostly priced between $20-$100. To figure out a name for the store, she hired a business consultant. The consultant came up with the name “Cena Luna”. What does that mean? I have no idea. But the name fits the vibe of the store perfectly. It’s also easy to spell, remember, and roll right off your tongue. Let’s see what ChatGPT can do:
Q: I want to open a small boutique with various gifts and trinkets in the $20-$100 dollar range in a small city. Give me a list of 10 names I could call the shop.
A:Certainly! Here are 10 name suggestions for your boutique:
Treasures & Trinkets Emporium
The Gift Nook Boutique
Whimsy & Wonder Gift Co.
The Charm Cabinet
The Curio Corner
Boutique Bliss: Gifts & Trinkets
Enchantment Emporium
Trinket Trove Boutique
Joyful Journeys: Gifts & More
Sparkle & Surprise Boutique
Feel free to mix and match or modify them to better suit your preferences and the vibe you want to convey for your boutique!
Ugh. Those names go from bland to terrible. No originality and nothing (at least in my opinion) near as good as Cena Luna. And this is coming from software that has access to every boutique name on the planet. Let’s give ChatGPT a little more prompting:
Q: My friend opened a boutique called Cena Luna that sold gifts in the $20-$100 range. What would some good alternative names be?
A: Certainly! Here are some alternative names for your friend's boutique:
Luna Belle Emporium
Celestial Treasures Co.
Moonlight Gifts & Trinkets
Luna Luxe Boutique
Starry Night Finds
Twilight Treasures Emporium
Crescent Charm Collective
Luna's Delightful Finds
Nightfall Nook Boutique
Midnight Sparkle Co.
These names maintain the celestial and elegant vibe of "Cena Luna" while offering a fresh twist.
Nope. It is interesting that even with the prompting, ChatGPT insists on adding a noun like emporium or co. to every suggestion. So yes, AI is going to change business. But beware of predictions that pronounce gloom and doom.