What Many CEOs and Management Teams Are Missing About AI

We’re at a moment where many CEOs are talking a lot about AI and there is generally an enormous amount of hype and media coverage about the topic.

Yet cutting through the hype, many (most?) CEOs and management teams are really tinkering with AI inside their own businesses -- they have yet to make any big bets using the technology, they have yet to scale AI solutions across entire departments, divisions, and enterprises. 

It is our view that tinkerers will lose in the new era of machine intelligence and (soon) superintelligence that is upon us. Here’s what many CEOs and management teams are missing: 

1. AI Is a Disruptive Platform Technology

From a business perspective, you can classify technology along two axes:

1. 1. How much it disrupts — Does it radically lower the cost of something valuable?
   2. How broadly it impacts — Does it affect one sector, or the entire economy and society?

Put that into a 2×2, and the top-right box – high disruption + economy-wide impact – is the rare “disruptive platform technology."

Exhibit 1 — Types of Technology Matrix:

Examples:

• Incremental Industry Innovation: Incrementally better saw blade in forestry industry — valuable for one sector, but no ripple effect elsewhere.
• Disruptive Industry Innovation: Hydraulic fracturing — slashed oil and gas extraction costs, but mostly stayed in energy. Another example here may be cloud computing (2006 onwards) which had most disruptive impact in the software and technology industry.
• Low Impact Platform Innovations: Email — sped up communications everywhere, but didn’t change cost structures radically.
• Disruptive Platform Technologies: Internet, AI — drove the cost of global communication, search, and distribution and human cognition toward zero, impacting every sector.

AI sits in that last category — and it’s bigger than any that came before.

2. The 9 Predecessors — and Why AI Is Different

AI is not the first disruptive platform technology. History gives us a clear playbook:

1. Printing Press (1450s) — Radically dropped the cost of mass information → literacy, science, reformations.
2. Steam Power (late 1700s) — Plummeted the cost of effective mechanical energy → factories, trains, ships.
3. Electricity (late 1800s) — Created very low cost power anywhere, anytime → productivity revolution in every sector.
4. Internal Combustion Engine (late 1800s–early 1900s) — Produced radically cheaper, mobile power → cars, planes, mechanized farming
5. Telecom (late 1800s) — Radically dropped the cost of real-time communication across continents.
6. Semiconductors & Computing (1950s onward) — Dropped the cost of computation exponentially → automation, finance, software.
7. The Internet (1990s) — Crushed the cost of global data distribution/search → near zero.
8. Mobile Revolution (2000s) — Radically lowered the cost of mobile computing and connectivity – put it in everyone’s pocket.
9. Social Platforms (2010s) — Radically lowered the cost of creating, maintaining, and scaling relationships globally.
10. AI / Machine Learning (2020s) — Now driving the cost of a wide range of cognitive tasks toward zero: content creation, analysis, design, decision-support, and pattern recognition. And in 2025, we’ve arguably crossed a historic threshold — the dawn of superintelligence in narrow domains, where AI can already outperform top human experts in reasoning, coding, research synthesis, and creative problem-solving. This is the first time in history that “better-than-human” cognition can be deployed instantly, at scale, for almost no marginal cost. 

Exhibit 2 — Major Disruptive Platform Technologies Timeline:

Why AI is different: Every prior platform technology accelerated one primary domain — energy, transportation, communication, or computation. AI touches all of them at once because cognition is the universal bottleneck in progress. And unlike electricity or steam, AI can self-improve, compounding its own capabilities at digital speed. That means the gap between early movers and laggards won’t just widen — it will become unbridgeable.

AI use generally is also happening with lightening speed. ChatGPT was launched in November 2022. In five days, it achieved 1 million users (a record for consumer applications). Just two months later, its user count soared to over 100 million and today it boasts over 800 million weekly average users.

3. History’s Lesson: All-In, At-Scale, Urgent

During the 2000 timeframe when the internet was cresting across every industry, we had front row seats to the disruption from different vantage points. We saw up close the bold actions that some companies took, and how others ran pilots and tinkered achieving little until it was too late. 

• Amazon — Bet the entire company on e-commerce infrastructure, scale, and speed; now one of the most valuable companies in history.
• Netflix — Shifted early from DVD rentals to streaming before broadband was ubiquitous.
• eBay — Rapidly built a global marketplace network while others were still testing “online classifieds.”

Others tinkered — treating the internet as a side project or a digital brochure — and paid the price:

• Blockbuster — Laughed off streaming until it was too late; filed for bankruptcy in 2010.
• Borders — Outsourced its online sales to Amazon in 2001 instead of building its own; collapsed a decade later.
• Kodak — Focused on film profits while ignoring the internet’s role in digital photography sharing and e-commerce.

The pattern repeats with every disruptive platform technology:

• Early movers compound advantage — talent, data, customer lock-in.
• Laggards can’t catch up, even with deep pockets, because the gap isn’t just tech — it’s capabilities, processes, and culture

4. What This Means for CEOs Now

If AI really is a disruptive platform technology (and it is), then tinkering is the riskiest choice you can make.

You need to:

• Go all in — Make AI a board-level and CEO-driven agenda item, not an IT side project.
• Act at scale — Embed AI in core products, processes, and decision-making across the enterprise.
• Focus on quality of models and outputs, not just adoption alone. It must be said that AI is also one of the only technologies in history to date that hallucinates (makes up information); and so care must be taken at present by CEOs to not just implement with speed but to focus on the quality of AI outputs and on establishing guardrails to reality test AI outputs.
• Move urgently — The compounding rate of AI progress is measured in months, not years.
• Understand this is all about culture at the end of the day. The Board and the CEO being “all in” on AI is a necessary but not sufficient condition. What is needed is full cultural change – and that requires full engagement with the entire organization from top to bottom. You have no hope of making real, lasting change unless the culture of the entire organization changes to one that embraces AI, sees it as a vital part of how every division, every department, and every team operates. 

The lesson from history is blunt: Tinkerers lose. In this wave, they won’t just lose market share — they’ll lose viability.

ABOUT THE AUTHOR(S)

John Kelleher is a Managing Partner and Global Head of CVC’s Operating team and is based in New York; Jean-Pierre Saad is a Managing Partner and Head of Technology Europe and is based in London; Sebastian Künne is a Partner and member of CVC’s Technology sector team and is based in London; Andrea Peyracchia is a Partner in CVC’s global Operating team and is based in Milan.