“The computer is incredibly fast, accurate, and stupid.
Man is unbelievably slow, inaccurate, and brilliant.
The marriage of the two is a challenge and opportunity beyond imagination.”
— Stuart G. Walesh (or Einstein)
Take the calculator. It is a marvellous piece of engineering, a machine that performs math faster and more accurately than any human ever could.
But it does not think. It does not understand.
When a button is pressed, it crunches the numbers. Feed it garbage, and that is exactly what it produces: perfectly calculated garbage. The calculator does not know what anyone wants. No intuition. No awareness. Just obedience.
That is what is happening with AI today.
It has been dressed up, given a voice, enabled to answer emails, write reports, and make decisions. But inside, it remains a machine. Fast. Confident. Clueless.
It has no understanding of businesses, customers, or aspirations. It simply processes. And now, entire companies, their products, and their strategies are being built around these machines. There is a danger in mistaking computation for comprehension.
This scenario has been seen before, with a different filter.
In the late ’90s, the internet promised to change everything, and it did. But first, it created a bubble. Startups with no real product reached billion-dollar valuations. Hype accelerated at warp speed. When the dust settled, trillions had vanished, not because the technology was wrong, but because the story raced ahead of reality.
Does this sound familiar?
Today, AI is climbing that same curve even faster. The hype is louder. The tools flashier. But the foundation is far more fragile. These systems do not understand the intricacies of any given world. They do not comprehend why things work. They merely remix patterns from the past. Yet, onto them, we entrust the keys to our organizations. This is not innovation; it is automation without insight.
The real danger lies here: when these systems fail, they do not pause or admit ignorance. They make confident mistakes at scale.
Statistics reveal that 88% of AI pilots never reach production. Of those that do, over 37% fail within the first year. Meanwhile, the gold rush is in full swing. Accenture, for example, rakes in billions, not from building models, but from advising others. Those selling shovels in this rush are the true winners.
This is not merely wasted money. It is an amplification of ignorance. Decision-making is increasingly entrusted to machines that sound intelligent, machines that give the impression of knowledge without real understanding. Noise is being scaled. Confusion industrialised.
This bubble will not burst simply due to inflated valuations. It will pop because its foundation is too thin. Castles cannot be built on sand without a solid ground beneath. When reality catches up, when returns fail to appear and strategies unravel, the crash will be intellectual as well as financial.
The world is not living through an era of artificial intelligence. It is living in an era of artificial obedience.
To change this trajectory, a demand for more is necessary. Progress requires building on truth, understanding, relentless curiosity, and human insight.
Without that, the future is not progress.
It is merely speeding into the fog.
And that is not the future anyone should want.
-G
Useful Links
RheoData: AI project failure rates ranging from 70% to 85%, reasons for failure, and strategies for success https://rheodata.com/ai-failures-stats/
LinkedIn article on why 75% of AI projects fail to deliver ROI, including Gartner and Accenture insights https://www.linkedin.com/pulse/why-75-ai-projects-fail-deliver-roiand-how-can-turn-things-minett-jssac
Cybersecurity Dive: Increasing share of businesses scrapping AI initiatives, with statistics from S&P Global https://www.cybersecuritydive.com/news/AI-project-fail-data-SPGlobal/742768/
S&P Global Market Intelligence: AI rapid adoption but elevated project failure rates and challenges https://www.spglobal.com/market-intelligence/en/news-insights/research/ai-experiences-rapid-adoption-but-with-mixed-outcomes-highlights-from-vote-ai-machine-learning
RAND Corporation research on why over 80% of AI projects fail and recommendations to succeed https://www.rand.org/pubs/research_reports/RRA2680-1.html
AIMultiple: Root causes of AI failures and real-life examples in 2025 https://research.aimultiple.com/ai-fail/
Informatica blog on why most AI projects fail and how to avoid it. https://www.informatica.com/blogs/the-surprising-reason-most-ai-projects-fail-and-how-to-avoid-it-at-your-enterprise.html
Gartner report forecasting high failure rate for agentic AI projects https://theaiinsider.tech/2025/06/26/gartner-forecasts-high-failure-rate-for-agentic-ai-projects-amid-rising-hype-and-cost/
Orion Data blog on beating the 80% data and AI project failure rate in 2025 https://www.oriondata.co.uk/post/unlock-success-how-to-beat-the-80-data-and-ai-project-failure-rate-in-2025