Commoditizing your complements: How Google, OpenAI, and China are playing different AI games

I paid $200/month for OpenAI’s Deep Research in February. By March, Google offered the same capability for free. This isn’t random—it’s strategic.

OpenAI and Google are playing different games. OpenAI monetizes directly, while Google protects its search business by making potential threats free. This follows Joel Spolsky’s “commoditize your complements” strategy: when complements get cheaper, demand for your core product rises.

It’s why Square gave away card readers (to sell payment processing), why Google invests in free internet access (to gain search users), and why Netscape gave away browsers (to sell servers). For Google, AI research tools are complements to search—making them free protects their primary revenue stream.

But China is playing an entirely different game. DeepSeek surprised Western researchers with its R1 model in January. Unlike Western companies focused on monetization, DeepSeek released their model with liberal open source licensing—unthinkable for Western AI labs.

The Chinese government designated DeepSeek a “national high-tech enterprise” with preferential treatment and subsidies. The Bank of China committed $137 billion to strengthen their AI supply chain, while provincial governments provide computing vouchers to AI startups.

This creates three distinct approaches:

• AI Startups (eg: OpenAI): Direct monetization of AI capabilities
• Tech Giants (eg: Google): Commoditization to protect core business
• China: National strategy for AI dominance without pressure for direct returns

What does this mean for AI development? Can Western startups survive when features are rapidly commoditized by tech giants while China pursues a national strategy? And which approach do you think will lead to the most significant AI advancements long-term?