The 151st issue of the Titanium Degree Picture
Those who win talents win the world.
With ChatGPT taking the world by storm, it has detonated a new round of AI craze. China and the U.S. are fighting for AI talent.
On April 17, according to the latest statistics from Titanium Media and Titanium Map, 32 Chinese and American technology giants, including Meta, Microsoft, Google, Amazon, Alibaba, and Tencent, have spent more than 5.7 billion yuan to recruit AI talents in the past year.
Among them, 16 U.S. tech giants spent as much as 5.63 billion yuan to win AI talent, accounting for more than 98% of the total spending, while China spent less than 100 million yuan.
Such a huge contrast in AI talent recruitment can be called "the first time in history". Even Tesla CEO Elon Musk (Elon Musk) bluntly said that this is the craziest AI talent "battle" he has ever seen.
So, what are the methods of major companies to poach people, and what is the flow of talents between companies?
The United States: 5.6 billion yuan to compete for AI talents, with the highest annual salary exceeding 18 million yuan
The first anniversary of ChatGPT's inception comes as the world experiences a year of explosion in AI applications. From intelligent writing assistants to image-based videos, AI launches an amazing version update almost every week.
Whether it is the industry represented by Silicon Valley technology giants such as OpenAI, Meta, and Google, or academic institutions such as Stanford University and the University of California, AI technology has become a key topic that cannot be avoided.
According to Amazon research, 73% of employers prioritize hiring AI talent, and 93% expect to use generative AI solutions in the next five years.
In fact, in the competition for AI talent, there is a consensus among major companies: "Those who win talent get the model, and those who get the model win the world." ”
On April 15 this year, the "2024 Artificial Intelligence Index Report" released by Li Feifei's team at the Institute for Human-Centered Artificial Intelligence (HAI) at Stanford University showed that in 2011, the proportion of new AI PhDs employed in industry (40.9%) and academia (41.6%) was roughly the same. However, in 2022, the proportion of PhDs entering industry after graduation (70.7%) was much larger than PhDs entering academia (20.0%) – and the proportion of PhDs in AI entering industry rose by 5.3 percentage points over the past year.
This shows that large-scale AI talent is moving from university campuses to industry.
According to the incomplete statistics of the salary of 16 technology companies, Silicon Valley technology giants represented by Meta, Microsoft, Google, Amazon, etc., have spent about 5.63 billion yuan to recruit top AI talents in the past year. And in this, huge sums of money play a key role.
In March, Inflection AI, the third-largest generative AI unicorn in the U.S. valued at more than $4 billion, announced that its co-founders, Mustafa Suleyman and Karen Simonyan, joined Microsoft as head of Microsoft's AI business and chief scientist, respectively. More than 70% (about 74%) of Inflection AI employees go to work at Microsoft because of high salaries and high positions.
It is reported that Microsoft paid $650 million to Inflection AI this time, of which $620 million was used to obtain the company's non-exclusive technology license, and the remaining $30 million was used to get Inflection AI to abandon its poaching behavior against Microsoft. As such, this is considered "Microsoft hollowed out Inflection AI".
From April 2022 to April 2023, basic salaries for managerial positions in AI and machine learning increased by 5% to 11%, while those for non-managerial positions increased by 13% to 19% over the same period, according to WTW's survey of more than 1,500 employers.
In fact, in the past 12 months, Meta has been the most generous "employer" engaged in AI positions, with an annual salary package of up to 18.155 million yuan for AI-related positions, and in addition to the basic salary, there are also generous stock and bonus incentives, of which Meta provides each employee with a total annual stock of up to 15.334 million yuan and a bonus of up to 723,000 yuan; stock awards and bonuses, etc.
Not only Meta, but OpenAI, which was named by Musk to "scatter coins", has a median total salary of $925,000, or about 6.69 million yuan, including stocks. Another data also shows that OpenAI pays some senior engineers a salary of $1.4 million (about 10.13 million yuan).
The Information reported that OpenAl promised an annual salary (mostly in the form of shares) of between $5 million and $10 million when poaching people from Google.
Zuhayeer Musa, co-founder of Levels.fyi, said that the six candidates who consulted about OpenAI job offers on the career service platform had an average salary of $925,000, including bonuses and equity. Meta's 344 machine learning and AI engineers earn a median salary of nearly $400,000 per year, including bonuses and equity.
However, it's clear that salary isn't the only criterion for these top AI talents. Computing power has also become one of the core attractions for talents who want to join AI companies.
Aravind Srinivas, the founder of Perplexity, revealed that he wanted to poach someone from Meta, but because there are not as many GPUs as Meta, the other party politely refused.
Previously, Zuckerberg revealed that by the end of 2024, Meta will have more than 350,000 Nvidia H100 chips.
In the final analysis, talents, capital, computing resources, and data resources are all key elements in the development of the AI large model industry.
Naveen Rao, head of AI at Databricks, said, "The talent we're looking for is changing dramatically. On the one hand, there is a surplus of technicians in low-demand positions, and on the other hand, there is a severe shortage of new top AI talent. ”
Domestic: Ali recruits the most AI talents, and ByteDance has the highest salary
Compared with Silicon Valley's AI mitigation, major domestic technology and Internet companies are also competing for AI talents with high salaries.
According to incomplete statistics from Titanium Media and Titanium Map, 16 technology companies represented by ByteDance, Tencent, Alibaba, and SenseTime have spent about 60.014 million yuan to recruit AI talents in the past year.
With the "Golden Three, Silver and Fourth" spring recruitment market, the demand for AI talent has surged. Related positions represented by AI and large models are a highlight and hot spot in the recruitment market this year.
According to the statistics of 11 technology companies by Titanium Media and Titanium Map, as of April 11, 2024, Alibaba has the largest number of newly recruited positions in AI technology, accounting for 10.1%, Tencent ranks second, with 5.9% of newly recruited positions related to AI technology, and Meituan and miHoYo are tied for third place, accounting for 4.7%.
In terms of funding, according to statistics, in the past year, 62.5% of domestic technology companies have given millions of annual salaries and high bonuses to attract top AI talents.
Take ByteDance, for example. According to public information, the total annual basic salary (1.527 million yuan) and average annual stock (1.57 million yuan) for AI-related positions within ByteDance exceed 3 million yuan, and the average annual bonus can reach up to 345,000 yuan, in addition, the total annual salary package of Meituan's AI talents is as high as 1.813 million yuan, second only to ByteDance.
According to public data, from 2021 to the present, miHoYo, NetEase, and Xiaohongshu have risen in more than three years, and the number of newly recruited AI technology-related positions has increased by about 2.5 percentage points.
In fact, China and the United States account for 65% of all AI talent, making them the world's largest sources and destinations for top AI talent.
According to a survey released by the MacroPolo think tank of the Paulson Institute, 70 percent of top AI talent currently works in institutions in China or the United States, and 65 percent of top AI talent comes from both China and the United States. At the same time, the United States is home to 60% of the world's top AI research institutions, and is the preferred employment destination for the world's most elite (top 2%) AI talents, and 57% of the most elite AI talents prefer to work in the United States. China is second only to the United States and is the biggest competitor of the United States, but only 12% of the most elite AI talent prefers to work in China, and the gap is very stark.
Since 2018, China has added more than 2,000 new AI-related undergraduate majors, of which more than 300 are at the most elite universities. At present, the scale of China's AI core industry has exceeded 500 billion yuan, and the number of enterprises has exceeded 4,300. It is predicted that in the future, the number of AI talents in China may surpass that of the United States.
Quan Qianhui, founder of the short video platform Laihua Technology, said that the Chinese population base is relatively large, it is easier to cultivate excellent talents, and the industry has forward-looking thinking and very high industry acumen. ChatGPT exploded last year, and a large part of ChatGPT's work is done by researchers educated in China. "We have been cultivating our own talents in China. ”
Talent is becoming the most valuable resource in the field of AI. AI talent is not only crucial to the growth of the company, but also related to the progress of the entire industry. As AI technology continues to advance, the competition for AI talent may become more fierce in the future.
(This article was first published on the Titanium Media App, drawing by Chu Yanmo, edited by Liu Yaning and Lin Zhijia, planning and produced by Titanium Media Visual Center)