Summary of the Key Points
This article completely breaks away from the extreme narratives prevalent in the market, which either glorify AI as a technology that will “rule the world” or criticize it as a mere scam designed to exploit consumers. Instead, it treats AI as a regular, emerging business sector, similar to e-commerce 20 years ago or electricity 100 years ago. There is no mystery around the technology; it’s all about tangible, measurable industrial logic. AI has just moved beyond the stage of investing heavily in technical prototypes and has entered a period of rapid growth, where it’s integrating into various industries to improve efficiency and generate real profits. All costs, benefits, risks, and rules are quantifiable, and the article provides practical guidelines for individuals, businesses, and investors, essentially lifting the veil of mystery surrounding AI.
---
Detailed Explanation of the Key Points
1. Demystifying AI: Its Real Motivation is to Help Businesses Save Money
Many people mistakenly think that AI is either meant to replace humans or is just a story created by internet companies to boost stock prices. However, research shows that while 86% of businesses are interested in AI, only 12% are eager to invest heavily in it. The majority of businesses see AI as a tool to improve efficiency or modify their existing processes. The logic behind their willingness to pay for AI services is simple: if investing $1 now can save $2 in the long run, they will do it. AI primarily replaces repetitive, time-consuming tasks that can be automated—for example, creating a PPT used to take more than two hours now can be done in 45 minutes with AI, potentially leading to a 40% reduction in the number of staff needed, resulting in direct cost savings that can be seen on the financial statements. In other words, AI is not some “black technology” but a tool for increasing efficiency, just like Excel or DingTalk before.
2. The Profit Logic of the AI Business Has Changed
In 2023, the industry was obsessed with developing large models, as they were seen as the key to success. But by 2024, the focus had shifted. Large models have become almost like commodities—available for a few dollars, with no real scarcity. The AI ecosystem can be divided into three layers: the upstream layer, which provides the computing power (chips and servers), the middle layer (the models themselves), and the downstream layer (the applications that use these models). Profits no longer stay with the model developers; they flow to the downstream applications. Successful AI companies focus on creating three types of products: tools for generating content automatically, intelligent equipment (such as robots in factories or delivery vehicles), and software that automates processes (e.g., expense reporting systems). The real opportunity lies in using AI to reinvent businesses related to digital content.
3. Most People Misunderstand the Economics of AI
Many assume that using AI will significantly reduce costs, but the calculation is more complex. The costs include:
- Training costs: Creating a large model is a one-time investment that will only get more expensive over time. Small companies cannot afford this.
- Inference costs: The cost of using AI services has dropped significantly, with GPU rental prices halving in recent years. Many basic AI services are now offered for free.
- Cost savings: Businesses that implement AI can see substantial savings, such as reducing labor costs by 60% in manufacturing or improving efficiency in offices.
- Overuse risks: While using AI can be cost-effective, excessive use can lead to increased overall expenses. For example, while individual AI calls may be cheaper, the cumulative cost over time can be significant.
4. Find Your Place in the AI Landscape
There’s no need for everyone to become an AI expert or start a company from scratch. Even if you’re in a different industry, using AI to improve efficiency can still benefit you. For example, using AI to streamline your work can increase your productivity and generate profits. The key is to find practical applications for AI that fit your business. China has a significant advantage in the manufacturing sector, with a large and cost-effective supply chain, which makes it ideal for implementing AI solutions.
5. The Challenges of AI Implementation
AI implementation faces practical challenges, such as limited chip production, power supply, and lack of high-quality data. These issues cannot be solved by simply investing more money. A more cautious approach is to start with small, repetitive tasks and gradually scale up. Additionally, AI’s growing energy consumption (3% of global electricity consumption, expected to rise to 15% by 2025) is a concern that affects all of us. The external costs of AI will be shared by society as a whole, including increased electricity costs and data privacy risks.
In summary, this article provides a realistic view of AI as a tangible, profitable business opportunity that requires a practical and cautious approach, rather than extreme optimism or fear.