第一财经

Doctor's Million-Dollar Annual Salary: Another Bubble in the Robotics Industry?

原文:博士的百万年薪,也是机器人的一种泡沫

Summary of Key Points

The robotics industry has been booming recently, with a surge of investment. In just June, several companies raised over 1 billion yuan in funding; Qianxun Intelligence gathered 4.5 billion yuan in three months, and Wujiedongli received more than 200 million US dollars. The money is primarily being invested in data, computing power, and talent. However, the industry also faces obvious bubbles, such as skyrocketing salaries for top talents—recently graduated PhDs can earn anywhere from 2 to 3 million yuan, with chief scientists commanding annual salaries of up to 124 million yuan. There are also issues like inflated company valuations and some robots that remain at a “demonstration stage” without practical applications. Industry insiders generally believe that these bubbles are a necessary part of the growth process for emerging industries. Just like the internet bubble, even if they burst, the technology, talent, and infrastructure developed will eventually create real value. Moreover, leading companies have ample cash reserves, so short-term research and development won’t be affected.

Detailed Analysis

1. Massive Investment in the Robotics Industry

This year, the robotics sector has seen even more funding than last year: Zixing Robot raised 1 billion yuan, Qianxun Intelligence accumulated 4.5 billion yuan in three months, and Wujiedongli received over 200 million US dollars. The money is being used in three main areas:

  • Data: To make robots more intelligent, they need to have access to a large amount of data, similar to how humans learn. Currently, robots have much less data than large language models like ChatGPT. Qianxun Intelligence has already collected 200,000 hours of real-world data and plans to reach the intelligence level of GPT3.5 by accumulating 1 million hours.
  • Computing Power: Training robot models requires powerful computing resources, which are very expensive (e.g., NVIDIA chips). Qianxun Intelligence is working with Diguarobot, a company that makes cleaning robot chips, to develop domestically produced high-performance chips.
  • Talent: Both data and computing power rely on skilled professionals, so companies are willing to pay high salaries to attract top talent.

2. Feisty Talent Competition and Rising Salaries

Robotic companies are competing fiercely for talent, driving up salaries:

  • High Salaries for Top Talent: Ubixion offers starting salaries of 15 million yuan for chief scientists in the embodied intelligence field, with some reaching up to 124 million yuan, requiring them to lead technical initiatives.
  • Value of Recent Graduates: Outstanding PhDs can receive offers of 2 to 3 million yuan, several times higher than what the internet industry offered them in 2015.
  • Shortage of Algorithm Engineers: According to Zhaopin data, there is a demand for 1.46 times more robot algorithm engineers than are available, with an average monthly salary of 23,000 yuan, which is considered high.
  • Source of Talent: Two-thirds of the talent comes from domestic universities (such as Tsinghua, Peking, Zhejiang Jiaotong), and one-third are returnees from abroad, further driving up salaries.

3. Are There Bubbles?

The industry acknowledges three main areas where bubbles exist:

  • Salary Bubbles: Talent salaries have increased by 50%-100% in the past two years, with PhD salaries doubling in just two to three years. Yu Hongxiang says, “These high-paying talents need to create viable business models; otherwise, it’s a bubble that won’t be sustainable.”
  • Inflated Valuations: Companies are willing to offer high salaries in hopes of future market growth, but many still don’t generate profits and rely on funding.
  • Lack of Practical Applications: Some robots are at a “demonstration” stage with low technical requirements and limited commercial potential.

However, everyone agrees that these bubbles are common in emerging industries (such as the internet and electric vehicles), representing bets by investors and industry players on the future.

4. Bubbles Are Not Necessarily Bad

Even with bubbles, the real value of the robotics industry is undeniable:

  • Positive Legacy: Just as the internet bubble burst in 2000, it led to lower infrastructure costs and the emergence of applications like WeChat and Taobao. The technology, core components, and production chains developed by high-paying talents will remain, creating real value as costs decrease.
  • Leading Companies Are Resilient: Top companies have enough cash to continue research and development despite market fluctuations.
  • Strong Demand: Zhaopin data shows a 57% year-on-year increase in demand for robot algorithm engineers, indicating sustained interest in the field.

5. Conclusion: A Turbulent Time of High Investment

The robotics industry is experiencing a period of heavy investment and emerging bubbles, but it’s also making real progress in technology. On one hand, funds are driving the accumulation of data, computing power, and talent, bringing robots closer to human-level intelligence. On the other hand, issues like inflated salaries and lack of practical applications need attention. However, these bubbles are a normal part of growth, and the essential technologies will ultimately transform our lives.

(The entire analysis explains the robotics industry’s funding, talent, challenges, and future prospects in plain language, making it easy for non-financial professionals to understand.)