Summary of the Core Content
This analysis essentially calculates the survival prospects of the telecommunications industry’s “biggest players” in the AI era. In the past, during the mobile internet era, telecom operators invested billions in laying fiber optic cables and building base stations. However, internet companies like WeChat, which relied on the telecom networks to generate revenue, stole billions in high-value revenues from services such as text messaging and voice calls. As a result, telecom operators could only earn meager fees for data usage, falling into a predicament where the more they invested in the network, the busier it became, yet their revenue remained stagnant. In the era of large-scale AI models, two main strategies have been debated in the industry: “using network advantages to extract protection fees from AI companies” or “investing heavily in developing their own general-purpose AI models to compete with companies like DouBao and DeepSeek for users.” Both approaches are unfeasible. The most practical option for telecom operators is to avoid direct competition with internet giants and leverage their unique national distribution networks, confidentiality and compliance qualifications, and the combination of network infrastructure and local computing power to provide AI services with added value. They must abandon the old model of charging based on the amount of AI usage, as otherwise, they will once again face a situation where demand for computing power surges while profits per unit of service decline, missing out on the entire benefits of the AI era.
---
Detailed and Easy-to-Understand Explanation
1. Understanding the Telecom Operators’ Initial “Tragedy”: Why OTT Is Such a Pain Point
Many people may not understand what OTT is, but simply put, it refers to internet companies that provide services to users without having to build their own networks or set up their own base stations, relying on the existing telecom infrastructure.
When the mobile internet first became popular, telecom operators thought they held a monopoly on the national network and could easily make huge profits. However, the emergence of WeChat completely eroded these profits. Previously, sending a text message cost 0.1 yuan, but now WeChat messages are free. The billions invested by telecom operators in building the network now only generate a few dozen yuan in monthly data usage fees. It’s like building a highway across the country, only for the valuable business conducted on it to go to the internet companies, with telecom operators earning a mere few yuan in fees. The more traffic there is, the lower the fee per user, and even though data usage has increased by 100 times, users’ monthly bills have barely gone up. This is the “OTT shadow” that remains deeply ingrained in telecom operators’ minds.
2. The Illusion of Using Network Advantages to “Strangle AI Companies”
Some suggest that telecom operators could use their network dominance to extract protection fees from AI companies. However, this is impossible for three key reasons:
- Regulatory Restrictions: Telecom networks are considered quasi-public infrastructure, and regulators prohibit them from using their networks to monopolize or discriminate against other companies.
- User Reactions: The existing mobile networks are already under heavy load, and any attempt by telecom operators to offer exclusive services to AI users would quickly lead to public backlash.
- Business Models: The user access points in the mobile internet era are already controlled by internet apps like WeChat and TikTok, so telecom operators have no direct access to users’ AI usage data. Any attempt to charge protection fees would be in vain, with only a few government and corporate定制 projects offering limited opportunities.
3. Investing Heavily in Developing General-Purpose AI Models
Another idea is for telecom operators to develop their own AI models to compete with companies like DouBao and DeepSeek. However, this is also unrealistic. The cost difference speaks for itself: while leading AI companies charge only 26 yuan for 120 million AI tokens, a telecom operator might charge 30 yuan for the same amount, costing five times more.
The reason for this is that telecom operators are not well-suited for developing general-purpose AI models. Their expertise lies in infrastructure management, not in AI product development or ecosystem management. Additionally, their pricing models are inherently more expensive due to higher costs associated with purchasing and operating hardware and managing licenses.
4. The Telecom Operators’ Unique Advantages
Telecom operators possess three unique advantages that no other company can match:
- Compliance for Government and Corporate Clients: Many AI projects require data to remain local and secure, and telecom operators have the necessary qualifications to meet these requirements, allowing them to secure high-value contracts.
- National Distribution Network: While companies like Alibaba Cloud and ByteDance can only serve a few major cities, telecom operators have a nationwide network of stores and teams that reach even the smallest communities.
- Integrated Network and Computing Solutions: They can offer integrated solutions that make it very difficult for other companies to switch providers.
5. The Real Barrier: Internal Institutional Barriers
The biggest obstacle is not external competition but the existing internal mechanisms that prevent telecom operators from reaping the benefits of the AI era. For decades, telecom operators have been evaluated based on separate KPIs, with each department working independently. Changing these mechanisms to encourage collaboration and share profits would be extremely challenging, as it would require significant changes to the company’s culture and internal structures.
In summary, telecom operators have unique opportunities to benefit from the AI era, but they must overcome internal barriers and adapt their business models to do so. Failing to do so will mean missing out on the entire potential of this transformative technology.