AI Companies Know Competition Is for Losers. They’re All Trying to Become the Monopoly in the Industry

“If you’re building a company, the goal is to create a monopoly and avoid competition. Competition is for losers,” PayPal co-founder Peter Thiel said in a 2014 talk at Stanford University. Interestingly, Thiel was introduced by OpenAI CEO Sam Altman.

Thiel, also a co-founder of Palantir, billionaire investor, and advocate for longevity research, emphasized a fundamental truth: All companies aspire to monopolies. They just don’t say it out loud to avoid legal scrutiny.

The technology world has already produced its fair share of monopolies: Windows dominates desktop operating systems, Android and iOS form a duopoly in mobile platforms, and Google holds a near-total monopoly in online search. Attempts to challenge these giants have largely failed. The competition was a losing game.

The AI Monopoly Race

The AI industry is in the midst of a similar battle, with companies competing fiercely to develop models and applications to beat the others and dominate the market.

Which model is truly superior? The latest release of Grok 3 has created buzz, but like its predecessors, it isn’t significantly better than competitors.

As AI models advance, they increasingly resemble each other—some excel in coding, others in writing, others in search. The differences are becoming marginal.

Improvements are getting smaller and more expensive. Grok 3 was trained on 100,000 Nvidia H100 GPUs, yet all it achieved was catching up to its rivals, not revolutionizing the field. GPT-4.5 is expected to follow the same pattern.

Since raw AI power is no longer a major differentiator, companies are shifting their focus:

1. Simplifying access to AI. OpenAI plans to streamline ChatGPT to reduce model complexity for users.
2. Creating new reasoning models and AI agents. Examples include DeepSeek R1, o3-mini, and Operator, designed to push users toward more expensive subscriptions.

The goal? To stand out from the competition and establish a monopoly.

In this race, there’s, for now, one leading player, at least in terms of the number of users. According to CNBC and data compiled by journalist Ed Zitron, here’s the estimated AI user base:

• ChatGPT: 400 million weekly active users.
• DeepSeek: 27 million monthly active users.
• Gemini: 18 million monthly active users.
• Copilot: 11 million monthly active users.
• Perplexity: 8 million monthly active users.
• Claude: 2 million monthly active users..

The numbers suggest that for millions, AI is synonymous with ChatGPT. OpenAI’s first-mover advantage has given it dominance—but will it last?

More from Xataka On

I’ve Tested Grok 3: It’s a Smart and Fast AI Model, but That’s No Longer Enough

The situation differs from past tech monopolies because OpenAI doesn’t control the platforms AI will run on. Consider the installed user base of tech giants:

• Windows: 1.4 billion active PCs.
• Android: 3.5 billion active devices
• iOS/iPadOS: 2.35 billion active devices.

Microsoft, Google, and Apple won’t sit back and let OpenAI take control. They will integrate their own AI into their ecosystems. This tension is already showing—Microsoft and OpenAI’s partnership is no longer as stable. Meanwhile, Google and Apple are gradually adding AI features to their operating systems.

These companies, already de facto monopolies in their respective markets, understand—just as Thiel famously stated—that competition is for losers. In fact, they may be even more transparent about it, especially since many of them once struggled against dominant players before securing their own monopolistic positions.

This pattern is evident across the digital economy. In audio streaming, Spotify holds an almost monopolistic grip with a 31.7% market share. In video streaming, the landscape is more fragmented, but Netflix remains a dominant force.

E-commerce is largely dictated by Amazon, while Facebook maintains an unshakable position in social media platforms. Other industries follow a similar trend, with Uber in ride-hailing, LinkedIn in professional networking, Match Group (Tinder, OKCupid) in online dating, and PayPal in digital payments—all serving as prime examples of market leaders that have effectively sidelined meaningful competition.

Competitors exist, but they rarely shift the balance. Firefox didn’t change browser dominance. Bing barely dented Google’s search empire. Alternative mobile OS options never seriously challenged Android and iOS.

Open Source AI: A Wild Card?

There’s one unpredictable element: open source AI models. Linux never conquered the desktop market, but it became the backbone of servers, cloud computing, and even mobile (via Android). Similarly, open source AI models could disrupt the AI race.

DeepSeek has made it clear that AI models are on track to become a commodity—something users will take for granted, much like having a smartphone in their pocket. The real differentiator won’t be the models themselves but the application layer—the services and tools that enable users to get the most out of AI systems.

This is where open source AI models are gaining ground. Companies like Perplexity and Freepik, despite not developing their own models, have leveraged open source alternatives to deliver powerful solutions. Their approach sends a clear message: the AI itself isn’t what matters most—it’s what you can do with it.

That might be the key. The company that delivers the best solutions will secure the ultimate prize in this industry. The goal isn’t just achieving AGI—though that’s certainly part of the equation—but rather using it as a means to a much bigger end.

The real ambition? Becoming the AI monopoly.

Image | Austris Augusts (Unsplash)

Related | The Other AI Race: China Rides the Wave of DeepSeek’s Viral Success