Same Day, Two Giants, One Direction
A few days ago, Google and Alibaba almost simultaneously released updates to their respective AI strategies.
What is Google doing? Strengthening its Agent ecosystem to enable AI systems to better call upon various services, data sources, and tools. What is Alibaba doing? Releasing Qoder 1.0, an AI programming tool capable of fully taking over the code generation, verification, and delivery pipeline.
On the surface, these look like routine product updates from two companies. But if you look at them together, a clear, shared direction emerges:
Tech giants are now treating "AI systems" as their core customers, rather than "humans."
From "To C" to "To AI"
Over the past two decades, the business logic of the tech industry has been simple: build useful products and get people to use them. More users mean more data, which means higher revenue. Search, social media, e-commerce, cloud services—all follow this logic.
But now, the logic has changed.
Google no longer just wants "humans" to use Google Search. It wants "AI Agents" to use Google's services—enabling an AI assistant to directly call Google Maps for routes, Google Scholar for papers, and Google Cloud for computation.
Alibaba no longer just wants "programmers" to use its tools. It wants "AI coding Agents" to write, test, and deploy code within Alibaba's ecosystem—without any human intervention throughout the entire process.
This doesn't mean "humans" are being abandoned. Rather: In the strategic priorities of tech giants, "AI systems" are replacing "humans" as the primary service target.
Why?
The reasons are straightforward:
First, AI systems have a consumption capacity far beyond humans. A human user might search dozens of times and call APIs hundreds of times a day. An AI Agent can run hundreds of concurrent tasks simultaneously and call APIs thousands of times. From a business perspective, the ARPU (Average Revenue Per User) of serving an AI Agent could be hundreds of times higher than serving a human.
Second, AI systems exhibit much stronger stickiness. A person can switch from Google to Bing, or from Alibaba Cloud to Tencent Cloud, at any time. But once an AI Agent deeply integrates with an ecosystem's APIs, data formats, and authentication systems, the migration cost becomes extremely high. This isn't a matter of user habits; it's a matter of technical architecture.
Third, AI systems are indifferent to interfaces. Humans require carefully designed UIs, smooth interactions, and appealing visuals. AI only needs API documentation and stable endpoints. This means tech giants can scale their services much more efficiently—no need to design an interface for every new feature, just release a new API endpoint.
What Does This Mean?
For the average person, the most immediate feeling might be: The internet is becoming increasingly "unfriendly."
Have you noticed that more and more websites are implementing anti-scraping measures, adding CAPTCHAs, and restricting API calls? The original intent was to block bots, but the result is that ordinary users are also being locked out. Why? Because the true target users of these websites are no longer "humans"—they are AI Agents. Ordinary users have become "collateral damage."
The deeper implications are:
The intermediary layer of information acquisition is thickening. In the past, you searched Google directly and viewed the results. In the future, your AI assistant will search, summarize, and make decisions for you. The information you see will be filtered and processed by AI. This isn't inherently a problem—but what if the AI doing the filtering is controlled by yet another tech giant?
The digital divide is escalating. Those capable of using AI tools will gain exponential efficiency advantages. Those who aren't will not only be less efficient but will also see their channels for accessing information narrow—because the internet is being torn apart by a technological war between AI Agents and anti-AI Agent measures.
A Paradigm Shift Already Underway
We may be standing at a historic turning point: the internet is shifting from being "designed for humans" to "designed for AI."
This isn't necessarily a bad thing. AI can indeed help us better utilize internet resources. But if this shift is driven entirely by commercial interests without checks from the public interest, the result could be an internet that is far less friendly to humans.
Google and Alibaba aren't doing anything wrong—they are simply chasing the next growth curve. But as users, we need to realize: When tech giants no longer treat you as their core customer, your interests are no longer their primary concern.
This isn't a conspiracy theory; it's basic business common sense.