Without a doubt, we are experiencing a massive wave triggered by the AI technology revolution, much like the internet and mobile internet before it. The launch of ChatGPT at the end of 2022 completely changed many user habits and sparked a productivity revolution across all industries. In the past two and a half years, OpenAI's valuation has skyrocketed from $14 billion to $300 billion. So, is OpenAI, as a leading player in the industry, truly worth $300 billion?

This question has sparked intense debate within our team:

Before we start, some basic information about OpenAI:

• OpenAI was founded in 2015 with the mission to advance research in Artificial General Intelligence (AGI).

• ChatGPT was launched on November 30, 2022, becoming one of the fastest-growing applications in history in terms of user acquisition.

• In a recent interview, Sam Altman stated that OpenAI's active users might have reached 1 billion. Public reports from earlier this year indicated a Monthly Active User (MAU) count of over 600 million.

• OpenAI's revenue primarily comes from subscriptions and API services. Currently, it has around 10-20 million consumer subscribers and approximately 3 million business users.

• Microsoft and SoftBank are OpenAI's largest external investors. Microsoft currently receives a 20% share of OpenAI's revenue.

Additionally, to provide a more intuitive understanding of what $300 billion and 1 billion users represent, here are some comparable products for reference:

It seems that $3 trillion is the current market valuation ceiling for a single tech sector. Nvidia, Bitcoin, and cloud services are all currently valued at roughly this scale.

Proponents' View: $300 Billion is Attainable

1. ChatGPT is the super-portal for AI, has already captured the mindshare of consumers, and its user base and stickiness continue to grow.

• Historically, consumer applications with over 1 billion active users have secured a firm place in the market and are difficult for latecomers to displace. Even if competitors have some product advantages, these tend to be neutralized during competition.

• ChatGPT has now become the default entry point for using AI products, potentially capturing the entire global knowledge worker population (about 1 billion people), making it very difficult for other companies, whether Anthropic or Google, to penetrate.

• Current monetization is still primarily based on subscriptions. There is vast potential to explore advertising, become the interface between humans and AI, act as a gateway for controlling robots, develop AI hardware, and more. The future potential is enormous.

2. In the mobile internet era, user value was between $100 - $200. ChatGPT's consumer user base alone could support a valuation of $100 - $200 billion. Let's make a detailed comparison:

• Regarding the value of OpenAI's consumer users, we make the following projection:

• This indicates that even with conservative adjustments, a $300 billion consumer-side valuation is very reasonable if the 1 billion DAU target is achieved and a fair portion of the per-user value of mature platforms is realized. (This consumer valuation is only part of OpenAI's potential total valuation; enterprise revenue, API services, and future innovations are expected to contribute even more value).

3. In terms of foundational large models, if it maintains its current development pace, it has a very high chance of becoming the AWS of LLMs.

• The market values AWS at around $1.5 trillion. Considering the future prospects of LLMs, the market size of LLMs is very likely to reach or even exceed that of cloud services. OpenAI is the undisputed leader in the LLM industry, and its market share could surpass AWS's—theoretically, the valuation ceiling for this part of the business could reach $1.5 - $2 trillion.

Conclusion: The future scale of the AI sector is promising. ChatGPT has a significant first-mover advantage. Considering its current user base of over 1 billion and its leading position in foundational models, a $300 billion valuation still has room to grow.

Opponents' View: $300 Billion is Too Expensive

1. Large model capabilities are becoming more commoditized, with diminishing differences between products, and low switching costs for both consumer and API users.

• ChatGPT defined the LLM space and the way users interact with large models (via chatbots), gaining a first-mover advantage. However, the capabilities of leading models are now quite close, and results from different models are similar in general-purpose scenarios. Some applications use a mix of underlying large models (e.g., Cursor, Notion AI). For B2B API users, the interchangeability of models is higher than that of cloud services (switching from AWS involves significant time and cost), perhaps more like a utility grid where it doesn't matter much which provider you use and you can switch at any time. For consumers, while there is some stickiness, migration is also very fast (compared to adding friends on social networks or training recommendation algorithms on content networks), and they can even use multiple services simultaneously.

2. The future of AI will be more tightly integrated with personal data and workflows. The current chatbot is just a transitional state, and whether OpenAI can successfully pivot has not been proven.

• The chatbot is a commodity form-factor—different users get the same answers. I believe the current product form is limited by the context length and token prices of previous models. With the continuous expansion of context windows and the rapid decrease in token prices (dropping more than 10x annually), there will be more AI Agent tools based on personal data as context. Theoretically, users won't need to leave their existing workflows (like using Gmail, Notion, or Slack) to get AI-assisted answers or even complete entire task flows (like an assistant or employee who can work continuously).

• Therefore, how many scenarios will still require us to open ChatGPT to interact with AI capabilities is an unknown.

3. Monetization:

• Currently, ChatGPT's primary revenue comes from subscriptions and API usage fees. Of course, with users, there's no shortage of business models to explore, but the process of exploring them is not without risk. These risk factors should also be comprehensively considered in the company's valuation.

Conclusion: ChatGPT is far ahead, but its current form may not be the ultimate form of AI interaction. The underlying model capabilities are increasingly on the path to commoditization, and competition is intensifying. Even if the future ceiling is $1 trillion, the numerous risk factors on the path to achieving it must be considered. A $300 billion valuation is still a bit expensive.