A decade ago, Carl Benedikt Frey and Michael A. Osborne made a startling prediction: Technology and automation would soon render up to half of all American jobs obsolete.

Their 2013 report, “The Future of Employment: How Susceptible Are Jobs to Computerization” was a doozy at the time.

The authors predicted that first on the chopping block would be lower-skill jobs that require little education and largely perform rote, repetitive administrative or technical tasks.

It turns out the they were not far off.

Just 10 years later, the World Economic Forum (WEF) published a Future of Jobs Report that estimated about a third of all business tasks are already performed by automation.

That is, most of the jobs Benedikt and Osborne were talking about have already vanished—or (one hopes) have been converted to other activities.

All that's needed are a comparatively few workers who know how to program AI bots and create prompts that leverage AI and machine learning technology to do many of these jobs for them.

The report also forecasted that in five years’ time, businesses would leverage AI applications to automate nearly half—47%—of all business procedures.

Since then, the WEF has reduced that estimate to about 42%. A bit lower, but still a colossal, cataclysmic economic event.

What is Generative AI?

Generative AI creates text, art, graphics, video, or other outputs that are nearly indistinguishable from human work—and in some cases better.

It’s a step above “Rudimentary AI”—which has been running customer service programs for years.

Every time you’re on the phone with your bank or your phone company and you get the automated message, “In a few words, state the reason for your call,” you’re engaging with a basic AI chatbot.

But that isn’t fooling anybody.

With generative AI, you’ll feel as you’re speaking with, reading the words of, or looking at a painting drawn by an actual human.

ChatGPT is today’s most well-known example of generative AI. But there are many others—some of which are working behind the scenes and are nearly undetectable as AI bots to the public.

For example, AI technology is already automating a lot of credit underwriting and credit card issuance—especially on the lower end of the credit market.

Banks and other financial services companies are also using AI bots to handle customer service tasks, lower their overall cost structures, and reduce overhead.

So far, AI has allowed all manner of banks and other firms with a big headcount of admin, clerical, and creative staff to become much more efficient.

The WEF projects that generative AI will create about 69 million jobs over the next five years, and render another 83 million jobs obsolete.

That’s a net job loss of about 26 million jobs.

A report from Accenture projects that AI technology could affect as much as 40% of all working hours, with estimates ranging from

9% to 63% of the typical workday

Overall, Goldman Sachs estimates that artificial intelligence could replace as many as 300 million jobs worldwide.

They also estimate that AI will increase the total annual value of goods and services sold by about 7%.

It’s not going to be just a few niche tech-focused companies, either.

AI is everywhere. The WEF’s 2023 Future of Jobs Report found that 75% of all organizations surveyed said they would be adopting AI within the next five years.

Sectors at risk

The cost: A massive and accelerating amount of what economist Joseph Schumpeter called, “creative destruction.”

As with every other major technological innovation, from the automobile to PCs to smartphones, this latest round of technological evolution will result in big winners and big losers.

A recent note to clients from UBS tries to answer that very question.

According to UBS analysts, generative AI’s biggest impact will be cost-cutting. And it will first hit media, software, insurance companies, and banks.

“US Bureau of Labor statistics suggest [that media, internet companies]— alongside banks and insurance companies—have the greatest proportion of their workforces exposed to potential automation,” wrote UBS analysts.

Eventually, the technology will penetrate other, less tech-savvy sectors—including luxury goods, medical devices, real estate, technology hardware, telecoms, and retail.

UBS analysts think the cost-cutting potential is so big that it will match that of past technological revolutions such as printing, electricity, or the railroad.

In fact, AI’s impact has already been felt in software development. Since the introduction of ChatGPT, visits to coding/developer forums have plummeted, according to UBS data:

visits to coding forums chart

Some sectors, however, are beyond the AI reach. UBS analysts think AI’s effects may be more limited for companies in construction, maintenance, and janitorial services.

For everyone else, the race to AI adoption is on.