Theme photo by BoliviaInteligente on Unsplash.
AI, AI, AI …
I don’t think I’m alone in saying I’m sick of hearing and reading about AI.
About how AI is awesome. About how AI is the future—and how the future is now. About how AI is going to take over all knowledge work and make tens of millions of people vocationally obsolete and irrelevant.
It’s hard not to be alarmed by the avalanche of depressing takes on LinkedIn about the AI revolution. Ten minutes of this doomscrolling is more than enough to make us increasingly anxious as we navigate today’s troubling world.
But, at some point, cooler heads need to prevail.
That time is now.
Especially for older workers wondering what it all means for them.
Let’s try to do that here. Let’s lay out the upside of AI, the downside, and the reality that lies somewhere in between.
Once we do, then we can talk about what late-career professionals can or should be doing now.
First, the good news …
Artificial intelligence marks a breathtaking technological step forward. But, what is it, exactly?
According to IBM, “Artificial intelligence (AI) is technology that enables computers and machines to simulate human learning, comprehension, problem solving, decision making, creativity, and autonomy.”
As for what, at a high level, AI can actually do, IBM also points out that, “Applications and devices equipped with AI can see and identify objects. They can understand and respond to human language.”
The uses for AI are many and are often profound. AI enables the analysis of large volumes of complex data, accelerates medical research, and predicts and corrects issues in machinery or processes before they arise.
Powerful stuff.
Different types of AI can be classified by technology, capability, functionality, and business purpose. But people have probably had the most exposure to generative AI, the predictive AI technology that can create images or videos.
Those crappy Coca-Cola holiday commercials we saw the past two years? Those were AI-generated using this technology.
So is a growing share of the pictures, video clips, memes, and text you see on social media. Much of this content you probably realize is AI-generated. Some (for better or worse) you may not recognize as the product of AI.
Yes, AI can do some amazing things. But, as the misuse (and sloppy use) of generative AI has made clear, it has a downside.
That, however, isn’t the worst of it.
Now, the bad news …
As AI grows more powerful and capable and enables individuals to quickly complete a wider array of tasks, companies will need fewer employees to get as much or more work done.
Some claim we are at or have recently passed an inflection point where most knowledge work—the work people typically do with their brains while sitting at a desk—can now be done by AI.
And as knowledge workers are encouraged or required to use AI tools, they help make the AI tools more powerful and better capable of replacing human roles.
At least initially, AI’s rise is predicted to have the greatest impact on young people, as companies will find that they don’t need to fill as many entry-level roles.
Older workers, meanwhile, especially those in the most vulnerable jobs (i.e., the ones AI can do more quickly and cheaply)—analysts, creators, accountants, attorneys, data scientists, customer service reps—also appear to be at risk of being displaced.
And as older workers tend to draw higher salaries and have a more difficult time finding a new job or switching to a new career, they would appear to have the most to lose in a mass AI-disruption.
Scary, right?
Well, before we allow ourselves to freak out too much, let’s stop right here and talk about … Garth Brooks.
The Garth Brooks Fallacy
At the height of his popularity in the 1990s, Garth Brooks gave an interview.
At that point, he was the biggest act in American music. He was ubiquitous on country radio. He had sold millions of albums and had seen several of his songs top the charts. He was playing sold-out shows in large arenas. He had reached a point where it was fair to ask what was possibly left for him to accomplish.
When asked some version of this question during the interview, Brooks responded something along the lines of, “There are 8 billion people on Earth. Do the math.”
What he meant was that there were still plenty of people for him to reach with his music and, therefore, there was still lots room left for his popularity to expand.
Only it didn’t. At the time of that interview, Brooks was at the pinnacle of his fame. In the years that followed, instead of becoming a bigger star, his popularity began to ebb. And while Brooks can still draw big crowds in sold-out arenas, it’s clear now that he’ll never come close to selling 8 billion records.
I think of Garth Brooks and his “do the math” hubris every time I see a proclamation about how AI will displace all knowledge workers, young and old alike.
Take, for instance, recent research by Anthropic, the company behind Claude. In this study, the company identified the proportion of job tasks in various fields that AI can replace. Then it offered an estimate for how much of that work is currently being done by AI. The gap between the two numbers represents the growth in AI usage that will happen, maybe sooner than later.
In other words, Anthropic, like Garth Brooks, did the math.
But unlike everybody’s onetime favorite Friend in Low Places, AI isn’t at the peak of its popularity. It isn’t a fad. It won’t fade away. It will only become more powerful, more prevalent.
That doesn’t mean, however, that it will fully consume all knowledge work, as many are predicting.
Barriers to AI growth
Here’s why I believe that.
At present, there are several factors that could impede or derail the inevitable rise of AI.
The major players are betting big.
In investment circles, there’s regular talk of an “AI bubble” that will, at some point, burst, just as the dot.com bubble burst at the beginning of this century.
Who knows if the Big Pin Prick (so to speak) is inevitable. But what’s undeniable is that the major AI players are all making huge bets on the future.
It’s also extremely unlikely that 1) all these bets will pay off, and 2) there won’t be some shakeout and consolidation.
For all of the blind optimism about the future of AI, there is no turning a blind eye to the reality that there will be winners and losers among the major players. It also means not all of the AI applications currently on the table will prove profitable, meaning the market could narrow just as easily as it could expand.
Many companies don’t yet understand how to use AI.
The AI companies aren’t the only ones betting big. Large corporations are, too.
But for all the money they are investing in AI, surprisingly few of these investments are strategic in nature.
Business leaders understand that AI is the future and that they need to rush to adopt it. But many don’t yet understand how AI can positively impact their businesses or, at a more basic level, how they should integrate it into their operations.
That doesn’t bode well for companies expecting to see an ROI on their AI dollars. The result is likely to be at least a short-term pullback in enterprise AI investment.
AI makes a lot of mistakes and will continue to do so.
Generative AI is prone to hallucinate or, to put it more accurately, just make stuff up.
Take the AI-generated summer reading list article that included multiple non-existent books and authors. The article appeared in multiple large newspapers in 2025 and proved to be a point of embarrassment for those publications.
And, many experts believe there will never be a way to fully prevent AI from hallucinating. The false outputs will continue to cause problems, including some that will inevitably prove exceedingly costly and dangerous.
The necessary infrastructure doesn’t yet exist.
More—thousands more—data centers are needed to accommodate the growing demand for AI. But the buildout is facing delays, growing opposition in many parts of the country, and the shortcomings of the electrical grid.
It will be many years before the needed infrastructure is in place. That amounts to a huge barrier to AI expansion.
Workers understand the implications, and they’re not playing along.
Despite encouragement and mandates from their employers, workers have been slow to adopt AI tools into their workflows.
There are multiple reasons for this.
For one, people are hesitant to embrace change and move away from routines and processes with which they have grown comfortable.
Also, employers have generally done a poor job of introducing AI tools into the workplace. Training has been lacking, as have assurances that AI can make workers more productive and valuable, as opposed to ultimately expendable.
And then there’s the fact that workers are onto the game. They understand that employers are perpetually seeking ways to reduce headcount or even eliminate entire job categories.
Why would workers want to help train AI tools that will make it easier for their employers to eventually not need them?
Sorting it all out
In other words, just as there were reasons Garth Brooks never sold 8 billion records and never will, there are reasons to doubt AI will ever reach its full, job-destroying potential.
That doesn’t mean, however, that AI is going away. Or that older workers can collectively stick their heads in the sand and not adapt to the AI-driven workplace.
In the next blog, we’ll talk about what older workers can and should do now.
