A Lesson From History: AI Is Not Likely To Kill Jobs

This is the published version of Forbes’ CEO newsletter, which offers the latest news for today’s and tomorrow’s business leaders and decision makers. Click here to get it delivered to your inbox every Sunday.

AI adoption in businesses is beginning to grow. According to McKinsey’s State of AI in Early 2024 survey, which was released last week, nearly three quarters of U.S. businesses are now using AI for at least one work function. And half are using it in two or more. It’s being used the most in marketing and sales organizations—34% of all use is in those types of companies—but it’s also popular in product or service development, IT and customer service.

At an event in New York last week, several McKinsey analysts talked about the survey and what the results mean. It’s early days for AI use now, and companies are still working on figuring out how to best benefit their business. Lareina Yee, senior partner and chair of McKinsey’s Technology Council, said that one of the most important metrics and markers when looking at how companies are using AI is whether the investments are tied to its critical business goals and framed as ways to accelerate value.

“I think you want to see companies that are trying to change work,” Yee said. “That means you have a technology architecture change, a process architecture change and a people architecture change.”

Alex Singla, senior partner and global leader of QuantumBlack, AI by McKinsey, said about 85% of dreamed up and tested AI use cases don’t get put into production. The hard part, he said, is finding the best uses and getting people to take advantage of it.

Erik Roth, senior partner and leader in McKinsey’s strategy practice, talked about the firm’s internal AI system known as Lilli, which brings together research documents and experts, instantly compiling relevant knowledge when an analyst is digging into a specific topic. The system is available to all consultants at this point, 72% of whom use it. The AI system is not only important for McKinsey consultants for research, but it also gives the firm a window into how to develop an AI system that gets adopted and does something uniquely powerful for their business—something clients often go to McKinsey to get advice about.

With Lilli internally—and with other AI use cases that clients see—Yee said AI can become less of a neat tech thing people are talking about and more of a solution worth adopting.

“I think there is an ‘aha moment,’ when it goes from, ‘This is a cool technology,’ to, ‘I could do it differently.’ …That ‘it’ could be whatever is the business priority,” Yee said. “I could actually capture data in my mainframe differently. Modernization of applications, I could actually rethink personalization for marketing. I could rethink workforce planning. So whatever that ‘aha’ is, it is an opportunity of possibility.”

But while AI can fundamentally change the way work is done, it does not have to be a job killer. In fact, it can do more to augment the current workforce and give them more job satisfaction. I talked to Heidi Messer, chairperson and cofounder of Collective[i], about how AI can help in that regard. An excerpt from our conversation is later in this newsletter.

ECONOMIC INDICATORS

No matter how much Americans feel like they are done with inflation, it continues to stick around. In April, headline inflation stayed stuck at 2.7%, according to the Commerce Department’s personal consumption expenditures index, released Friday. This index, which measures how much more Americans spent in aggregate on goods and services compared to a year before, was also 2.7% in March. Core inflation on this metric, which excludes more volatile food and energy prices, was 2.8%. Inflation has been leading many consumers to spend less in recent months—something that was directly reflected in last Thursday’s downward revision of first quarter GDP, from 1.6% to 1.3%.

Basically, those who are holding out for rate cuts may find themselves waiting a bit longer. The news did little to move markets. The S&P 500 and the Dow Jones Industrial Average inch up a bit on Friday, while the Nasdaq index fell about a tenth of a percent. However, Forbes’ Derek Saul writes, the S&P 500 had its best May since 2009, and the second-best May in the last 20 years. The biggest reason for that rally were huge gains from most of the high tech “magnificent seven” stocks—Microsoft, Apple, Meta, Nvidia, Alphabet and Amazon. (The seventh of that group, Tesla, fell only 1% last month, but is down more than 28% so far in 2024.)

BIG MOVES

It seems the term “Big Oil” is getting to be outdated. As the recent spate of mergers and acquisitions in the industry has shown, perhaps it should be known as “Bigger Oil”—as of last week, another oil giant is set to get bigger. ConocoPhillips announced its plan to buy Marathon Oil in an all-stock deal for $22.5 billion. Forbes senior contributor David Blackmon writes this merger would create a mega-company largely focused in the Eagle Ford Shale and Permian Basin in Texas and New Mexico, and the Bakken Shale region spanning North Dakota and Montana. Many analysts are positive about the deal—as long as it is approved by the FTC—though Forbes senior contributor Gaurav Sharma points out a central piece of the strategy behind the deal: The two oil companies are combining as a matter of survival. The deal seems to be more about consolidation than growth, as many companies and nations seek to transition to more sustainable energy sources.

Compared with other recent oil industry deals, this one is actually pretty small. Earlier this month, ExxonMobil closed on its $60 billion all-stock purchase of Pioneer Natural Resource. Chevron is getting close to completing its $54 billion purchase of Hess, whose shareholders approved it last week. All three of these deals serve to give companies a bigger piece of the pie—and potentially better returns even as oil begins to fall out of favor.

NOTABLE NEWS

After losing a proxy battle in April to get on the board of Disney, activist investor Nelson Peltz sold his entire stake in the media and entertainment company last week. CNBC reported he made about $1 billion on the sale. Peltz, owner of Trian Fund Management, was trying to get himself and former Disney CFO Jay Rasulo on the board in an effort to force Disney to take a more urgent approach to its poor earnings growth and return on investment. Neither Peltz nor Disney responded to Forbes’ request for comments, but in April, Peltz told CNBC that if Disney was able to make the turnaround it promised, the company wouldn’t be hearing more from him.

Disney’s stock has been slowly trending downwards since the board vote in April, but it hasn’t fallen too far or too quickly. Forbes senior contributor Jim Osman breaks down some of the big challenges Disney is facing—a to-do list that’s incumbent on CEO Bob Iger to solve. Iger has an extremely long track record at Disney. He first became CEO in 2005 and retired from the job in 2021. He returned in November 2022, and plans to serve in the position until 2026. Iger has made some of the most transformative changes and acquisitions in the company’s history, Osman notes, but faces new challenges trying to make streaming services reliable revenue generators, keeping the company’s theme parks profitable, and responding to socially conservative critics of the company’s DEI positioning. For Disney’s continued success, Osman writes, Iger needs to be able to pivot from the strategic insight he brought to the beginning of his tenure to flexibility around the needs of a changing consumer.

TOMORROW’S TRENDS

Collective[i]’s Heidi Messer On How AI Will Help Workers

AI is going to revolutionize the workplace, but not in the way that some people think. Collective[i] helps businesses use AI to improve productivity. I talked to cofounder and chairperson Heidi Messer about how AI can save jobs and help employees. This conversation has been edited for length, clarity and continuity.

People are always saying AI is going to kill jobs. Will it?

Messer: Like any broad statement, there’s elements of truth to it, and elements of things that are not accurate or lead you to the wrong conclusion. If you look at history, every technological revolution has created more jobs than it’s destroyed. I think the argument that there will be net fewer jobs at the end of this current technological revolution is not supported by historical precedent. If you look at what AI does, some of the jobs that AI has the potential to eliminate—or I would say fundamentally change to the point where the people who are in them don’t have the skills to continue doing the job in its new form—are jobs that don’t leverage the full capacity of human skills.

In AI, we’re trying to replicate the greatest data processor that’s ever been invented in history, which is the human brain. The fact of the matter is what we can do on our best day is emulate portions of that brain that are the least sophisticated. For example, an LLM replicating language, being able to predict the next word that someone’s going to say, is not a super-complicated function that our brain performs. It’s awe-inspiring to us because we very rarely see how the sausage is made, and we see it happen with OpenAI, it’s incredibly exciting. But the reality is, an LLM isn’t going to come up with the next theory of relativity. It’s just not an innovator. It’s not a creative thinker. It can make probabilistic predictions that cover a large swath of the way communication is done, and it can replicate a poorly-written content blog, but it’s not going to write a Pulitzer Prize-winning novel about a new subject. And so the things that AI is designed to do are commoditized brain functions. Things that require large swaths of data and training data to be able to come up with answers, but aren’t like innovative human thinking.

So where does that leave jobs? For this, again, I would go back to history. If you were, at the beginning of the Industrial Revolution, the best horse and buggy driver that ever existed, knowing what was coming down the pike from cars, your job would change. You’d have to learn how to drive a car instead of a horse and buggy.

The same is true for AI. I think that people are going to have to acquire a whole new set of skills to be competitive in the industry. And right now we have this very weird situation of underemployment. If you talk to any AI company in the world, if you talk to any big corporation, they’re scrambling for talent: data science talent. They’re scrambling for engineering talent. They’re scrambling for highly creative talent. That’s because they don’t need the low-level mass content bloggers. They don’t need the people who were doing assembly line functions that were repetitive, and candidly, not that thought-provoking.

What about employee retention and job creation? What can AI do along those lines?

AI should remove tasks that people hate to do. For example, one of the things we do is automate data entry into CRM. Nobody likes being a data entry clerk. Today, for most sales teams as an example, that’s 20% of their job. The best salespeople just don’t do it, which means their companies have no record of the work that they’re doing, and no way to replicate things that are being successful. But everybody’s being forced to do something they hate to do.

If you can start to remove tasks that people hate doing and replace those with much higher-value tasks or—and this is where the people side of it comes in—training, you can hone human skills. One of the things that hopefully the AI era will usher in is a mass upskilling, because you’re moving away from a world of mastery into a world of continuous learning. That means that companies have to start being able to support that continuous learning in order to make AI successful as a companion to people.

So on its best day, AI is a companion. It’s a polymath that people have in their pocket that helps them make better decisions. It does the grunt work that nobody wants to do, and it makes people more successful, elevating the skills that make us all unique.

Companies are just beginning to figure out how to make AI work for them. Five years in the future, it will presumably be more integrated and a more normal part of working. At that point, how do you see AI’s role with the workforce?

I think it’s a collaboration. We can look at some examples of where this has already happened. Would you ever get on a plane where the pilot made an announcement and said, ‘I’ve shut off autopilot because I think I have better judgment?’ But we didn’t get rid of pilots, right? We have pilots who supervise the AI.

I think there’ll be some combination of AI working in isolation on tasks that are completely automated. I mentioned the example of automating data entry into CRM. You don’t need a human being to supervise that. The AI could do it. You’ll have some instances where AI is collaborating with people. For example, we have in beta a chatbot that is a polymath for senior executives to learn about their businesses, and be able to talk to an AI and get advice. And then in some cases, you’ll see AI that works for people. That would be similar to the autopilot example, where a pilot is supervising AI and essentially being a check on its mistakes, but also taking it to another level.

FACTS + COMMENTS

Amazon announced Friday it was adding food delivery from Grubhub Plus to its benefits for users with Prime memberships.

$10: Monthly cost for a Grubhub Plus subscription, which includes free delivery for orders $12 and up

180 million: Amazon Prime subscribers in the U.S. as of March, according to Consumer Intelligence Research Partners

‘Bring more convenience to Amazon customers’: What Grubhub CEO Howard Migdal said the Prime deal does

STRATEGIES + ADVICE

Generative AI is not magical by any means, but it can have a transformative impact on your business. Here are some things you need to understand about the reality of AI before you can implement it.

Businesses that give back tend to reap many benefits from doing something good. Here’s why you should make it a priority.

VIDEO

Miley Cyrus On Balancing Her Business Endeavors And Creativity

QUIZ

Billboard Boxscore reported that concert tours between last October and March grossed an unprecedented $1.5 billion. Total receipts are much more than that because one act with a hot show didn’t report its gross totals. Which is it?

A. Taylor Swift

B. Madonna

C. Pink

D. U2

See if you got the answer right here.