$NASDAQ(.IXIC)$  THE GREAT AI CIRCUS OF 2026

Everybody wants Artificial Intelligence. Unfortunately, nobody seems particularly interested in Actual Intelligence.

Every CEO wants AI. Every board wants AI. Every consulting firm has suddenly become an AI expert. Every technology vendor has discovered that adding AI to an existing product makes it worth three times as much. And every company seems to have developed the same revolutionary strategy: Announce an AI transformation, retrench 20% of the workforce, hire a Chief AI Officer, ask the remaining employees to figure out what AI is supposed to do, and issue a press release celebrating a 40% improvement in productivity. Welcome to the Great AI Revolution, where companies are firing the people who understand the business and replacing them with technology that needs those very people to explain the business.

1. The CEO Who Discovered AI Last Tuesday

A CEO attends a technology conference. Someone wearing expensive sneakers announces that AI will transform every industry on the planet. The CEO hears phrases like Generative AI, Agentic AI, Autonomous Enterprise and Tenfold Productivity. He understands approximately 7% of what is being said. But he understands one thing perfectly: HIS COMPETITORS ARE TALKING ABOUT AI! By Monday morning, the executive committee has been summoned. "WE NEED AN AI STRATEGY!" The CIO asks, "What business problem are we trying to solve?" The CEO looks horrified. "Business problem? EVERYBODY IS DOING AI!" The CFO immediately sees an opportunity. "If AI can do the work of ten people, perhaps we can get rid of eight." Nobody asks whether AI can actually do the work. Those are technical details. The CEO has a shareholder presentation next Thursday.

2. Congratulations! You Have Retrenched Your Company's Intelligence.

A company has been operating for 30 years. Its experienced employees understand the products, customers, systems, business rules, exceptions and undocumented processes. They know why a strange rule was introduced in 2009 and why the production system doesn't always behave according to the documentation. They know where the bodies are buried. Not literally, although given some corporate IT projects, perhaps we shouldn't investigate too closely. Then management announces that AI will enable a leaner workforce. Out go the experienced employees. Three months later, the AI team asks, "Who understands this product?" Silence. "Where are the business rules?" More silence. "Who designed this interface?" Retrenched. "Who knows the exceptions?" Early retirement. Congratulations! The company has successfully eliminated the human intelligence required to implement its artificial intelligence. The CFO calls it efficiency. The remaining employees call it Friday.

3. The Magical Prompt: Apparently Anyone Can Ask AI Anything

"Just ask ChatGPT!" Wonderful advice. Ask it what? Someone who doesn't understand a subject often doesn't know what questions to ask. Imagine a junior employee asking AI how to configure interest calculations for a banking product. AI produces five beautiful paragraphs, tables, formulas and a professional summary. Fantastic! Except the explanation may be based on generic banking concepts rather than the bank's actual product configuration. It might be 90% correct and 10% catastrophically wrong. A senior specialist would recognise the problem immediately. A junior employee might copy the answer into a design specification. Somebody implements it. Six months later, 80,000 accounts have incorrect interest calculations. Management announces, "We have identified an opportunity to strengthen our AI governance framework." Translation: NOBODY CHECKED WHETHER THE BLOODY ANSWER WAS RIGHT.

4. ChatGPT Knows Your Company's Secret Products? Really?

A CEO announces, "We will use ChatGPT to transform our internal operations." Excellent. Ask ChatGPT for the exact business rules governing your company's proprietary Product XYZ. If those rules were never published, documented or made accessible to the model, how would it know? Telepathy? Corporate clairvoyance? A psychic connection to your internal servers? General-purpose AI models do not automatically possess your company's confidential specifications or undocumented business rules. Companies can connect AI to internal documents and databases without training their own models, but the information must exist and be reliable. Now imagine the company has outdated documentation, disconnected databases and historical decisions buried in emails. Management has just retrenched the people who understood everything. But don't worry. They purchased 5,000 AI licences. Problem solved!

5. The $2 Million Discovery of Missing Knowledge

CEO: "Why isn't our AI answering product questions correctly?" Consultant: "Your documentation is incomplete." CEO: "Can AI complete it?" Consultant: "We need reliable source information." CEO: "From whom?" Consultant: "Your product specialists." CEO: "We retrenched them." Consultant: "What about your historical documentation?" CEO: "Outdated." Consultant: "Your operational procedures?" CEO: "Nobody maintains them." Consultant: "Excellent. I recommend a six-month AI Readiness Assessment." Fee: $2 million. The CEO approves immediately. Finally, some progress.

6. The Wonderful Mathematics of AI Productivity

Company announces, "AI has improved productivity by 40%!" Fantastic. How was that measured? Did revenue increase? Did costs fall after accounting for licences, infrastructure, integration and verification? Or did someone demonstrate that AI could generate a 20-page report in 30 seconds? If an employee previously spent three hours preparing a report and AI now generates it in two minutes, wonderful. But if the employee then spends four hours correcting errors, checking figures and removing imaginary references, congratulations: productivity has actually declined. Naturally, the presentation highlights the two-minute generation time. The four hours of checking become "human-in-the-loop value enhancement." Beautiful terminology. Absolutely magnificent.

7. The Great Retrenchment Paradox

AI genuinely can automate tasks, accelerate development and improve efficiency. Some jobs will change, and some organisations may legitimately need fewer employees. But the experienced specialists who understand business rules and exceptions are often the people best equipped to implement AI successfully. Unfortunately, they are expensive. So out they go. Six months later, the company hires external consultants at three times their previous rates to rediscover the knowledge those employees already possessed. The consultants recommend workshops, documentation exercises and specialist recruitment. The new job advertisement reads: "Seeking industry professional with 20 years' domain experience to support AI transformation." Salary: 30% higher than the person they retrenched. You have to admire the efficiency.

8. The Board Has Discovered a New Language

A normal search function becomes "AI-Powered Knowledge Discovery." A chatbot becomes "Intelligent Conversational Engagement." An automated workflow becomes "Agentic Enterprise Orchestration." A spreadsheet with formulas becomes "Advanced Algorithmic Decision Support." And if a product has absolutely nothing to do with AI? Add an AI button. Preferably blue. Investors love blue buttons. The next quarterly presentation proudly announces, "We have embedded artificial intelligence into 17 strategic business processes." Translation: "We added a chatbot to the intranet, and three employees have used it." One asked about annual leave. One asked for the cafeteria menu. The third asked whether the CEO's job could be automated. That employee is now attending a mandatory HR discussion.

9. The Question Nobody Wants to Ask: Who Is Accountable?

Suppose AI generates an incorrect financial report, recommends a faulty loan decision or modifies production software incorrectly. Who is accountable? The AI? Good luck bringing a language model into a disciplinary hearing. The vendor? The employee? The CIO? The CEO who announced that AI would eliminate human supervision? Suddenly, everybody rediscovers the importance of human oversight. "We must keep a human in the loop!" Excellent. Which human? The expert you retrenched in March? Or the remaining junior employee expected to review 600 AI-generated decisions before lunchtime? That's not oversight. That's decorative supervision.

10. The Year Is 2028. The AI Transformation Is Complete.

CEO: "Ladies and gentlemen, our AI transformation is complete. We have reduced headcount by 45%!" Board: "OUTSTANDING!" CEO: "AI now handles product queries, customer operations, documentation and internal decision support." Board: "REMARKABLE!" Director: "And are the answers correct?" CEO: "Excellent question." Director: "Who verifies them?" CEO: "Our experienced employees." HR Director: "We retrenched them." CEO: "Then ask AI to verify itself." CIO: "It says everything is correct." Board: "FANTASTIC!" The share price rises 12%. The CEO receives a $15 million performance bonus. Three months later, a customer discovers that AI has been providing incorrect product information for six months. An emergency board meeting is called. CEO: "We have identified a strategic opportunity to reintegrate human expertise into our AI operating model." Translation: "DOES ANYBODY STILL HAVE THE TELEPHONE NUMBER OF THE PEOPLE WE FIRED?" Unfortunately, they're now working for competitors. At higher salaries. Using AI. Properly.

Final Thought

AI is genuinely powerful technology with enormous potential. But buying AI is not the same as understanding AI, generating answers is not the same as producing truth, and reducing headcount is not necessarily improving productivity. Somewhere in a corporate boardroom, a CEO who has never personally tested AI against his company's actual products is explaining how it will replace half his workforce. His presentation is magnificent. Because ChatGPT wrote it. And nobody checked it.

Welcome to the Artificial Intelligence Revolution — where the machines are getting smarter, the PowerPoint slides are getting prettier, and the people making the decisions are still trying to work out what the question was. 😂

- Written by yours truly ...AI