Way to data scientist

Google I/O 2025: From ‘Moat Crisis’ to the ‘Quadripartite’ AI Empire Counterattack

Google I/O 2025 Conference: A Real AI Comprehensive Transformation

this year’s Google I/O conference can be said to be the most shocking in recent years. The most striking, undoubtedly, is the release of The Veo 3 video generation model. This model marks a leap in video generation technology: from “only generating images” to “truly understanding and generating dynamic video content”, it opens up a huge imagination for the application of AI in video creation, advertising, film and television and other fields.

Not only that, Google also released a complete AI “family bucket” at this conference, covering all levels from basic models, tool chains to terminal applications, demonstrating its determination and strength to go to war in the era of large models.

Never been optimistic about the “full stack self-research” industry moat.

In the past period of time, Google has been considered the least favored among the “Seven Sisters of US stocks” because its search business is facing unprecedented challenges:

• in Q1 2025, the volume of Google search queries on Apple devices declined for the first time, and the volume of Google search in Safari browser fell for the first time in 20 years;
• Gartner predicts that by 2026, traffic to traditional search engines will be reduced by 25% as users turn to AI tools;
• the market has doubts about its growth prospects, so it is given the lowest price-to-earnings ratio valuation in the group.

However, Google did not stand still, but took a series of positive transformation measures:

• launch a new AI-driven search model to completely reshape the traditional search experience;
• explore advertising embedding possibilities in the conversational experience of generative AI;
• last year, I/O launched the “AI brief introduction” function, which has been implemented for 1.5 billion months;
• Google emphasized that “overall search query volume is still growing, and the total number of requests from Apple platforms is also rising”.

These actions show that Google is completing a refactoring of the underlying logic, rather than a simple “plug-in” of AI functions “.

Google’s comprehensive restructuring strategy: not only AI, but also Platform Revolution

the significance of this transformation is comparable to Microsoft’s transformation from selling licenses to selling cloud services. Google is trying to shift from an advertising-based business model to a new business system centered on member subscriptions, API services, and AI terminal tools.

This refactoring path is fundamentally different from other AI players:

• OpenAI have a strong model, but lack of terminal platform;
• Meta has the flow entrance, but in the computing power and chip level is still controlled by people;
• microsoft excellent integration capabilities, but Ecology, terminals and chips all rely on external partners.

Google is currently the only player to realize the four-in-one capability of “model + chip + computing power infrastructure + Global Terminal entrance”, which constitutes its real moat.

Key metrics to focus on:

the success of Google’s AI transformation depends not only on its technological leadership, but also on its breakthrough in product commercialization. There are a few key milestones to follow:

• Gemini Ultra user growth and ARPU (average revenue per user) for subscription services;
• progress in the integration of AI search models and advertising systems;
• The penetration rate and retention rate of AI products (such as Veo and Gemini App) on the end side;
• API call volume and conversion rate of enterprise-class AI capability.

Write at the end

it is not difficult to see that Google’s giant ship has begun to turn around and evolve towards “AI native platform. If this transition goes through, it will not only open up a whole new growth curve for Google, but it could also redefine the way people use search, content generation and AI tools.

So, It’s time to pay close attention to Google’s stock