Using AI to destroy your business

Jun 22, 2025

You decided to read this. That means you think AI is an efficient way to destroy your business. You're right. AI destroys businesses faster than slow decline.

Follow these steps to transform your company into a failure case study.

1. Abandon Everything for AI

a corporate boardroom with executives frantically throwing profitable business plans into a trash bin labeled "OLD BUSINESS" while chasing after floating holographic AI symbols. Revenue charts showing steady profits lie abandoned on the floor, covered in dust. In the background, a "CORE BUSINESS" office sits empty with cobwebs, while everyone crowds around a glowing "AI FUTURE" banner. The CEO holds a shovel, literally burying traditional business models in a corporate cemetery

You have a core business that pays salaries and thrived for years. But it's boring. It lacks AI. AI is the future.

Let your past go. Focus only on the future.

Invest everything in AI. Ignore existing revenue streams. Let your core business drift. It survived this long—it can survive alone. Chase every AI fad instead.

Your customers will understand. Your services may deteriorate and your product may suffer, but it has AI. They'll appreciate that.

2. Divide and Lose

an office floor split into dozens of tiny, walled cubicles, each with a different AI project banner - "AI Chatbot Team," "Generative Art Squad," "Random Algorithm Group." Workers in each section wear noise-canceling headphones, completely ignoring each other. Arrows showing communication flow all point inward to each silo, with no connections between groups. In the center, a confused customer stands alone, completely ignored, holding a "HELP" sign while everyone types furiously on their own AI projects.

Split your company into thousands of mini organizations. Give each its own agenda. Create power struggles.

Everyone attends every AI conference. Nobody talks to each other. Most importantly, nobody talks to customers about their needs.

Each group builds its own AI solution. Your company should create useless agents, generative art plugins, and random chatbots. None should relate to your core business.

3. Gut Your Core Operations for AI Talent

A successful factory floor or server room being stripped bare as talented engineers carry their tools toward a gleaming "AI DEPARTMENT" sign. Behind them, critical systems spark and smoke with "OUT OF ORDER" signs. A manager with a clipboard checks off names: "Best Engineer ✓ Moved to AI," "Top Developer ✓ Moved to AI." Meanwhile, alarms flash red in the abandoned operations center, and money literally flows out of broken pipes labeled "REVENUE STREAMS."

Pull every quality engineer from core business operations. Watch something mysterious happen: unmitigated disaster.

Move all important talent from revenue-generating streams to chase the latest vendor offerings. Let your infrastructure crumble without talent and experience.

When questioned, remind everyone: "Transformation comes with pain."

4. Reward Wishful Thinking Over Results

an awards ceremony where employees receive giant trophies for "BIGGEST PROMISE" and "MOST OPTIMISTIC PROJECTION" while actual working prototypes sit in a trash bin labeled "BORING RESULTS." A scoreboard shows "PROMISE SIZE: 10/10" next to "ACTUAL DELIVERY: 0/10." In the audience, one person holding a small but functional AI tool looks dejected while others cheer for elaborate PowerPoint presentations of impossible AI fantasies floating as holograms above the stage.

Create incentives that celebrate big promises over delivered results. Judge employees on promise size, not achievements. Reward those who paint rosy AI pictures. Silence those who focus on practical problems and AI reality.

Value wishful thinking. Celebrate ridiculous infrastructure costs and technical debt accumulation. If someone proposes incremental implementation, remind them: "Big companies don't think small."

5. Choose Tools Over Processes

a chaotic office where every desk is buried under stacks of different AI tools, software boxes, and evaluation frameworks. Workers frantically juggle multiple laptops running different AI platforms while completely ignoring a simple, effective workflow chart crumpled in the corner. A massive, incomprehensible dashboard covers an entire wall with thousands of meaningless blinking metrics, while a simple "ACTUAL PROBLEMS" whiteboard sits blank and ignored in the background.

Obsess over evaluation frameworks that apply to all problems yet effectively solve none.

Always throw tools at problems. Ignore processes entirely. Incentivize people to drop the latest tool on every issue. Tech debt doesn’t matter.

Make dashboards so complicated they drown you in meaningless metrics. Never know actual system problems—make up answers instead. Your MBA prepared you for this.

Use "LLM as judge" for evaluating business outcomes. If an LLM says you're right, who are you to argue? You're just an algorithm's humble servant.

6. Never Look at Your Data

a executive in an expensive suit sitting on a throne made of unexamined data printouts and hard drives, wearing a blindfold labeled "INTUITION ONLY." Around him, charts and graphs showing actual business insights float ignored in the air. A junior analyst in the corner tries to present real findings but is waved away. Meanwhile, the executive consults a Magic 8-Ball labeled "ML DEGREE" for business decisions, while spreadsheets with real solutions burn in a fireplace marked "TOO SIMPLE."

Data analysis is for junior analysts. Don't waste time applying domain expertise to your data. What could you find that someone with a fancy ML degree and no business sense wouldn't?

Trust your intuition. If it looks good, it is good. When someone produces counter-evidence, don't give them time. Remind them it's beneath you to examine data.

If you must look at data, buy the most complex analytical tool available. Simple Excel spreadsheets don't suit you.

7. Turn Everyone Into Jargon Joe

A meeting room where every person speaks in word bubbles filled with incomprehensible technical jargon: "NEUROMORPHIC BLOCKCHAIN SYNERGY," "QUANTUM ML PARADIGMS," "AGILE AI VECTORIZATION." Above their heads, confused question marks float while a translation device shows "MEANING: UNKNOWN." One person in the corner speaks in plain English ("Let's solve the customer's problem") but their words appear as tiny, ignored whispers. A "COMPLEXITY METER" on the wall shows maximum levels.

Don't listen to people who know. Listen to people who drop random jargon into conversations. Every meeting should become techno-babble where everyone stays confused because nobody understands anything.

Make obfuscation your competitive advantage. Clients should never understand your proposals. Business teams shouldn't know what they're participating in. If someone without a computer science degree understands you, you're failing.

Hide behind complexity. Simple isn't impressive.

8. Reduce AI to a Buzzword

P.S. This post uses inversion, based on discussions with friends, colleagues, and many AI gurus especially Greg & Hamel. To succeed with AI, do the exact opposite of everything above.

AI should answer every business problem, operational issue, and need. If anyone does anything besides AI, they do nothing.

Improve margins? AI will do it.

Convert leads? AI will do it.

Retain talent? AI will do it.

Create a business strategy so lengthy and complicated that nobody reads it. Yet your strategy should focus only two letters: AI.

Never diagnose business problems.

Never talk about pragmatic solutions.

Never coordinate strategic efforts.

Be all over the place, because AI can be everywhere. When questioned, tell them to get up to speed on AI and read your lengthy strategy document.

The End Result

Follow these prescriptions and you'll confuse motion for progress, tools for strategy, and complexity for competence.

Your competitors will thank you for the learning experience. Your employees will find better companies. Your customers will discover better products and services.

But who cares if you lack good service or products? You have AI demos to show.

P.S. This post uses inversion. It is based on discussions with friends, colleagues, and many AI gurus - especially Greg & Hamel. To succeed with AI, do the exact opposite of everything above.