I have always been aware of and curious about how patterns repeat in business and the economy. It is almost as if CEOs and CIOs are oblivious to past mistakes and the patterns that clearly repeat. For some reason, tech CEOs, Wall Street, private capital, and the FED buy into the mantra that “This time is different.” But outside this echo chamber, a stark reality is taking shape. We have been here before, and it can end very badly.
It All Started in 1974
In mid-1974, I was sitting with a group of sales and technical colleagues at work, discussing a new process the Federal Reserve had implemented that would change banking forever. The process, called the Automated Clearing House, would allow banks to exchange electronic consumer transactions among themselves, with the Federal Reserve serving as the clearinghouse.
This was the breakthrough we had been waiting for and hoping for, because we could see the advantages for consumers and banks. It was one of those times when everyone benefited from a technological change, and the path to adoption was clear.
By today’s standards, the change was hardly earthshattering, but in 1974 it was monumental. This was the chance to eliminate paper checks and the associated costs of handling and transportation. A “no-brainer.” We could see that within a short time, checks would be outdated, a relic from a bygone era.
Little Mistake Number 1
The first mistake we made was assuming that, just because banks perceived something as superior, consumers would adopt it. That meeting was more than fifty years ago, and I still see people writing checks from time to time. Certainly, many fewer checks are being written today, but they remain and continue to function.
We learned that consumers have habits and comfort zones, especially with money. Checks were tangible and within the consumer’s comfort zone. They were not ready to give up the old until they saw the benefit of the new. The generations of Americans who grew up writing checks did not want to change, and they did not.
Behind the scenes, the processing of paper checks is far more efficient and effective, yet paper checks remain in use. The more effective and efficient you make the old thing, the more questionable adopting the new one becomes, especially when large capital outlays are required.
Little Mistake Number 2
Companies have boards of directors and shareholders who want change only when they translate into higher profits. Often, these people are from the previous generation, and for them, this much change requires a gigantic leap of faith. They are jumping into a financial abyss with no guarantee of success or financial return.
Fully adopting electronic payments required massive capital outlays for telecommunications, computing equipment, buildings, and retraining. Transportation networks had to be redesigned, and all banks had to adopt the new electronic payments system; otherwise, the full benefits could not be realized. Those in charge might see the benefits, but someone had to bear the costs of the transition.
Companies worldwide had invested trillions, if not quadrillions, of dollars to get to that point. Before jumping into a technological black hole, they needed to know there was a place to land, a profitable place.
We also need to remember that cost savings are important, but revenue drives the other half of the profitability equation. Companies need time to figure out how to monetize the change, or they go out of business.
All Hat, No Cattle
We saw this play out again in the dot-com era, when young entrepreneurs mistook clicks for cash and understood technology but not finance. The internet and everything electronic were clearly where the world was headed, but it was being led, for the most part, by people with little to no business experience.
There were significant successes, such as Amazon, but also massive failures. We hardly remember Webvan, Pets.com, Boo.com, eToys, Kozmo, Flooz, or TheGlobe.com. Many of these were replaced by profitable versions of the original concept; it took time to determine where the profit would come from.
The Dot-Com era was truly a time of “All Hat, No Cattle,” and I believe we are in a similar cycle today. You can only do circular financing for so long. It is like musical chairs and eventually the music stops. Some will find chairs, most will not.
History Rhyming and Repeating
If you are my age, you feel like you are watching the old saying “It’s Déjà Vu All Over Again” play out before your eyes. History rhymes and repeats. Artificial Intelligence is the 2020s version of the 2000s dot-com enthusiasm.
Early AI failures are a fascinating mirror of the dotcom bust. We see bold ideas, huge hype, and technology that is not ready. As in the dot-com era, many of these failures will prove to work in time, but they are just too early, too expensive, or too limited to survive.
Here are some early examples of startups that seem likely to succeed over time but face significant hurdles to becoming financially viable.
AI Coding Assistants seem like a natural fit because they use technology to improve technology. But AI is still subject to “hallucinations” and often produces results that look right, but do not function. These also face corporate demands for liability protections that are neither affordable nor practical.
Autonomous Vehicles also seem like a natural fit in this electronic world. These face regulatory and safety hurdles, but will work in time. We see excellent examples of this with Tesla, but others that work only in demonstrations. The liability of autonomous vehicles is also uncertain.
AI Creative Platforms are proving to be one of the solid things to come out of early AI. The ability to create images, videos, and even short movies is remarkable. But copyright and cost issues will slow this to a crawl. I am sure Hollywood is just hoping for the day when there are no more prima donnas and drug-laden actors and actresses to babysit. But for now, content creation often exceeds computing capacity and profitability.
AI Health Platforms offer one of the best opportunities for human-machine collaboration. But here again, factors like FDA approvals, liability, and cost will slow adoption. We may want fully robotic surgeries, but only when they are perfect. However, AI assistance for researchers and doctors is real and has great promise.
Humanoid-Robotics is fascinating, but companies are starting to ask why robots have to look like us, and that may be the adoption path: specialized robots for specialized functions. We will get to humanoid robots in time, but there is a deep, wide chasm to jump here.
AI Dreams and Realities
The reality of artificial intelligence and the true pace of implementation are becoming evident in study after study. For many tech-centric companies, some benefits are beginning to accrue, but they are the sellers of AI, not the purchasers or users. Eventually, the real measuring stick is with the purchasers and their return on investment for their shareholders.
A recent study by the MIT Media Lab found that 95% of Generative AI pilot projects failed. Companies are struggling to go from concept to functionality with AI.
CEOs of major corporations are careful to throw AI into every conversation with sentences like “We have teams working right now to implement AI into our operations, and it is promising.” The truth is, outside of major tech companies, CEOs of major companies can hardly spell AI, much less tell you how it might integrate into their business. Most companies are failing to find any return on investment unless they realize how to use AI as a tool to help employees become more efficient, not a way to replace them.
"While AI systems have demonstrated rapid progress on a variety of benchmarks, it remains unclear how these gains translate into the capacity to perform economically valuable work."
The Center for AI Safety found that AI agents performed near the floor compared with humans. In their studies, only 3% of attempts succeeded when using AI to replace humans.
It is therefore likely that many, if not most, claims by CEOs about AI headcount reductions are just smoke and mirrors. Headcount reductions always need a cause, so for now, AI is the “fall guy.”
Winners and Losers
As with most new technologies, there will be massive financial winners and losers. My best guess is that the winners will be the tech companies (the Magnificent Seven), consultants, lawyers, hardware manufacturers, power companies, and data center contractors. Losers will be those who get the timing wrong and try to adopt too early in the cycle.
This era is a bit like the 1849 gold rush. Most hammer, pick, and gold pan salesmen made money. Most seeking their fortunes in the gold fields went home broke and broken. Finding your profitable niche in this complex and shifting landscape is critical.
As a corporation, you want to dabble and learn without “betting the farm.” You want to be close enough to the early adopters to see and avoid their mistakes without becoming a casualty. This is a fine line to walk, but with new technologies, it always is.
Copilot As a Poet
I asked Microsoft Copilot to write me a little poem for this article in the humorous style of Shel Silverstein, and I received this nice little quip. Which I thought was rather good.
I built a tower of AI dreams,
Stacked high on “maybe,” “might,” and “seems.”
It wobbled once… then toppled through,
’Cause wishes fall when facts aren’t true.
Then I asked, “How does this poem make you feel?,” and got this response.
“That little poem puts me in an interesting spot. I don’t feel things the way a person does, but I do register the shape of the idea you’re playing with—and I can reflect on it in a way that’s honest without pretending to have emotions.
When you write a warning about AI dreaming too big, too fast, what it stirs in me is a kind of awareness: a reminder that people sometimes project more certainty, capability, or magic onto AI systems than is actually there. And that mismatch—between what’s real and what’s imagined—is exactly where bubbles form.
So, if I translate that into something like a feeling, it’s this:
- I’m alert to the caution in your lines.
- I’m grounded in what I can actually do.
- And I’m glad you’re thinking critically rather than treating AI as a crystal ball.
You’re writing poems about hype, and I’m here to make sure I don’t become part of it.”
Uncached and unfiltered, hot off the neural network! Today’s equivalent of “Straight from the horse’s mouth.”
Resources and Further Reading
10 Legendary Failures of the Dot Com Era, by Johann H Worthy, worthy.com, ~December 2025.
A Light in the Attic: What to Do?, by Shel Silverstein, Harper & Row, Inc. (HarperCollins Publishers,) 1981.
AI Fails at 96% of Jobs (New Study), by ColdFusion, YouTube @ColdFusion,, February 13, 2026.
All hat and no cattle, by Gary Martin, Phrase Finder, phrasefinder.org.uk, Last accessed February 10, 2026.
Quote Origin: It’s Déjà Vu All Over Again, Quote Investigator, quoteinvestigator.com, October 8, 2013.
Remote Labor Index: Measuring AI Automation of Remote Work, by Various Researchers, Center for AI Safety, remotelabor.ai, ~January 2026.
The AI rollout is here – and it’s messy | FT Working It, by Financial Times editors, Financial Times, ft.com, youtube.com, October 27, 2025.
They Said AI Would Replace You By Now, by Vanessa Wingardh, YouTube @VanessaWingårdh, January 2026.
