I once shared this viewpoint:

Whenever a new technology rises and flourishes, opening a new era, although old powers and old players may still hold a place in this new era, those in a leading and monopolistic position are most likely new forces and new strengths.

A recent sign in the field of AI has increasingly shown this characteristic.

Just last year, among the large model companies in the United States, four competitors were strongly favored: OpenAI, Anthropic, Gemini, and Grok.

Among these four, except for Gemini, the other three are all new forces in the field of AI.

Although Grok is a new force, it has fallen behind earlier due to various reasons.

Among the remaining three competitors, Gemini, backed by Google, was once highly anticipated by the industry.

Because Google is not only the inventor of the Transformer, but also possesses top-tier AI companies like DeepMind in the research field. Moreover, Google has made significant financial investments in AI, with even the founders personally leading AI development.

However, under such circumstances, the latest version of Gemini has been delayed.

This puts Google in a very disadvantageous position in the currently heated competitive landscape.

As closed-source large models have developed to this point, especially when Anthropic's large model is already able to self-evolve and self-iterate, if Gemini cannot keep up, the gap between it and Anthropic will widen at an exponential rate.

Therefore, it is highly likely that in the competition of closed-source large models in the United States, the only contenders left who can take the lead are basically the completely new forces of Anthropic and OpenAI—existing competitors are being eliminated one by one, and future competitors will also find it difficult to seize the headlines in the field of closed-source large models.

All along, many people have held quite optimistic estimates of Google's business, believing that Google not only dominates the traditional search field but also has a high probability of becoming a leading competitor in the AI era, continuing to expand its business territory in this new era.

But in reality, this may be precisely Google's awkward point:

A significant portion of its business revenue comes from search advertising. In the AI era, due to the complete transformation of the Q&A mode, the search business will inevitably be increasingly diminished, and the resulting advertising revenue will also suffer severe impact.

On one hand, increasingly advanced AI will challenge Google's core business; on the other hand, existing competitive habits will push Google to find ways to preserve advertising, which is entirely like fighting against oneself.

In Jin Yong's martial arts novels, fighting against oneself can lead to extraordinary martial arts skills, becoming number one in the world. But in real life, fighting against oneself only leads to either exhausting oneself or defeating oneself.

In the competition of large models in the United States, Google is increasingly showing this awkwardness; similarly, in the competition of large models in China, Alibaba is also increasingly demonstrating this awkwardness.

I originally had high hopes for Alibaba and its developed large model business, but after paying attention for a period of time, I increasingly felt that something was off, somewhat like AI for the sake of AI, without clear direction on how to develop AI, despite Thousand Questions' remarkable performance in the open-source field and Alibaba's substantial investment and strong determination in large models.

Recently, after watching an interview with Cai Chongxin about AI, I felt even stronger about Alibaba's ambiguity and lack of clear direction.

So I estimate that in the competition of large models in China, the likely top player will also be the new forces and new strengths led by DeepSeek.

This does not mean that in the end, only a few players will remain in the AI large model space in both the United States and China, but rather that in the end, only a few players may hold an absolute monopolistic position, while other players will only share the remaining market share.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。