The most expensive start-up: artificial intelligence grab people battle

On the day before the Spring Festival, Qi Wanli was suddenly curious and wanted to see how many job postings were received in the mailbox since January 2016.

The final figure scared him - more than 700, equivalent to two emails a day from various headhunters. Digging his text messages, telephone calls, and emails, “Sometimes I receive several ones a day. At the beginning, I also meticulously replied. I was really bored afterwards.”

Qian Wanli is an artificial intelligence scientist of Alibaba Cloud. Like him, scientists engaged in artificial intelligence research are now ready to live. Even the middle-level and younger employees of his men’s team were on the headhunting list. From April last year, a group of young technical people who had graduated for two or three years had also been dug up.

The recruitment season for recent graduates is even more an empty match between companies. Not just money, the founder's halo, invitations for candidates to visit, etc. All the cards that can be played are thrown.

One morning after the three rounds of interviews, Tsinghua Computer's graduate student Du Xiao talked with Li Kaifu in the dorm room to chat with WeChat. "The conversation with Kai Fu was particularly funny, because some candidates did not believe that he could really talk to him. At the beginning, he also made a speech in order to prove his identity."

Last fall, after the establishment of the Innovation Factory AI Engineering Institute, it has been recruiting on a large scale. In order to rush to the Academy, Kellie, the person in charge of HR, mobilized Li Kaifu to personally communicate with key candidates. Last year he personally ran several college lectures.

"It is difficult for people to recruit. Even if they are graduate candidates, they can all be communicated several times before they are finalized. Now it's nothing at all for San Mao." Kellie said, this is the market this year.

In order to rob people, the company went a year before the students graduated. “We started with the standard to do a screening at a relatively high altitude, but we also saw that some companies returned to the entire package and slowly eliminated it.” Ali Yun’s head of human resources, Yu Wei, told 36.

According to the data that Boss hired directly to 36, the data service industry related to artificial intelligence is the fastest-growing area for the company's salary budget this year, with an increase of 41.9% compared with the same period of last year.

Data from Boss Direct Employment

The AI ​​is getting hotter and the pay rises, and behind it is a huge cliff for talent supply and demand – According to Alex Ren, founder of AI recruitment service TalentSeer, the supply and demand ratio for the posts related to artificial intelligence development in North America was nearly one to three. The situation is even more extreme, even closer to 1:10.

There are large companies such as large foreign companies, BAT, Huawei, and Lenovo, as well as startup companies. Moreover, everyone's observation is that start-up companies are particularly "out of the price".

Mobile Internet bonuses have reached the top. With the exception of live broadcasts and shared bicycles, there has been no new “venture” in the field of venture capital investment in the recent year. Artificial intelligence is almost the only one remaining.

In the field of artificial intelligence, talent is the core asset. This industry is now on the eve of breaking the technological bottleneck and putting it into use. No one has talked about anything.

Today, AI talent is in a high-price era of extreme imbalance between supply and demand.

The graduates in artificial intelligence related to high-quality colleges and universities do not lack the olive branch that comes out this year.

According to a graduate from the computer department of a certain line of colleges and universities, “25w per year is not too high, around 30w is normal, and foreign companies are higher, 35w to 40w. Startup companies are generally around 30w, and there are options.”

Roger, the founder of Internet search company Imatch talent, was quite impressed. When he graduated from Peking University's Department of Intelligent Science ten years ago, the monthly salary on the market was 8,000 to 10,000, even the most advanced companies like Microsoft Research Asia. The annual salary is only 12w to 15w. In contrast, today's AI pays for graduates, even if they do not count stocks and options, cash has also risen two or three times.

For more than ten years, the attitude of the talent market for artificial intelligence-related graduates has turned 180 degrees.

Wu Jie, a senior technical manager of the innovation workshop AI Institute of Engineering, was afraid to say that his major was "artificial intelligence and pattern recognition" when he graduated from the job search in 2004. "I said I was worried that I couldn't find a job. At that time, the fire was corporate computing, Java. ".

But now, according to a technical person in a startup company in the AI ​​field, “Engineers who can graduate in three years and can play with deep learning go to the headlines today, and it is normal for the annual salary to be 70w to 80w.”

The same is true of senior headhunter Chen Jing's observation: After working for two years at BAT, jumping to a startup company, you can often get 50-60w annual salary. This is just not including the cash portion of the option.

This year's Spring Festival, the above-mentioned technical person communicates with the classmate who is serving at BAT. The other party joked that if you want to dig me, at least 1000w of cash can dig me. "You see, taking into account pay and stocks, his level of people is already worth it."

No matter whether it is a graduate, a engineer with years of experience, or an advanced talent, the price is not expensive.

“The cost of labor is too high, and the company is struggling,” an industry source told 36. Many companies in the robot circle have already planned to recruit for second-tier cities like Changsha, Chengdu, Qingdao and Xi’an.

The high salary has caused many entrepreneurs to start eating too much. A consensus in the industry is that today in Beijing and Shenzhen, software-based AI startups, employee compensation has accounted for 50% of total spending.

Even more troubling is that there is no money to dug people.

Looking for AI talent has become a "looking" good business.

Usually, once the candidate recommended by the headhunter is hired, the headhunter can obtain the amount of 20% of the candidate's annual salary as a share. According to Chen Jing, a veteran headhunter who has worked for more than five years in the middle and senior-level talents, the average industry earns about 890,000 yuan. However, if he is a single AI or algorithm candidate, he can get more than 100,000 or even two. One hundred thousand.

However, Chen Jing’s team decided this year not to focus on AI digging, “not worth it”.

The demand-side market is very hot. Internet companies on the B-round and above require algorithm talents. However, because the supply is too small, headhunters have fierce competition in the field of artificial intelligence. On average, each candidate has more than a dozen headhunters staring and digging people. The half-year cycle has been considered very short.

Last year, Chen Jing's team did not successfully deliver a single AI business. "It's very hard for a candidate in the AI ​​field to get it. Half of the candidates in other industries have a 60% chance of rejecting us on the phone as soon as they come up, but the AI ​​and algorithm people's rejection rate is 99%," he said.

There is a huge supply shortage of AI technical talent.

The first reason is that there are gaps in the training and market demand of colleges and universities. Domestic colleges and universities usually go to postgraduate stage, will open artificial intelligence, algorithm-related courses. This means that the annual output of qualified personnel is very limited. Each year, there are about 20-30 postgraduates in the field of artificial intelligence and data algorithms related to a school, with about 10 doctors.

This is a basic technology industry that is now considered to be "water" and "air." Such a supply is tantamount to a drop in the bucket.

And the plates available for selection are not large. From Roger's point of view, fewer than 20 high-quality graduates who can truly provide artificial intelligence are included in such institutions and research institutes as Tsinghua University, Peking University, Shanghai Jiaotong University, Harbin Institute of Technology and Chinese Academy of Sciences.

Whether or not people can be dug up and who can be dug has become a key factor in the outcome of AI competition.

Today, AI-related and fall-placed applications, among the two most successful companies, Zhang Yiming, the founder of today's headline, graduated from software engineering, worked at Microsoft, participated in Cool News, and was a technical talent; RenHuaHua, who dropped out of school during Tsinghua's reading of Bo, joined Google China to study the application of machine learning in search.

In other words, whether or not the founding team can know people and whether they can dig people is crucial.

The law of digging people in the AI ​​field is that the top AI technical talents are often passive job seekers, seeking job opportunities through the relationships between teachers and colleagues, or accepting direct invitations from company founders. They are often buried deep within the company and can be said to be hidden in the water.

"In fact, people who are really busy doing things are not the type of industry opinion leaders in the media that have a high exposure. Therefore, the headhunters may not be able to dig them up and they need to look for new ways." Qian Wanli told 36.

Two of the country's most important technology companies are digging into the “stricken areas”, Microsoft Asia Research Institute and Baidu. They have cultivated a large number of research talents who have mastered advanced technology. However, so far, a large number of internal talents have been attracted by outside entrepreneurial opportunities.

A headhunter revealed that “Observing Microsoft’s tightest is the headline today.” Just a month ago, the billion-dollar unicorn just succeeded in digging into the former vice president of Microsoft Research Asia, Ma Weiying. After joining today's headline, Ma Weiying will serve as vice president of the company, responsible for managing today's headline artificial intelligence laboratory. In addition, Ding Haifeng, the founder of the core recommendation system architecture and infrastructure of today’s headlines, is now a senior engineer of the company's data department. He previously came from Baidu’s Web search department.

The flow of technical talent is often a series of walks. Usually after the departure of an AI project leader, his mentor will also follow him to the new team. One took five and it is not uncommon.

“This is an intellectual attraction. Just like martial arts world, you have a high martial arts person who is willing to mix with you.” Wu Jie said that because engineers must always cooperate with others and cannot fight alone, so often a boss can bring A team came out.

Domestic talents are expensive and the number is extremely scarce. Some startup companies that completed B-round financing and whose valuation is more than a few hundred million US dollars simply decided to directly dig people overseas. In the most recent half of the year, TalentSeer, an AI recruitment service company founded by Alex Ren, received a total of 200 jobs in more than a dozen companies in the AI ​​field.

How much does it cost to build an AI technical team in Silicon Valley? Alex made a calculation for 36 points: a team's organizational structure will show a pyramid structure. The top is usually a Ph.D., graduated more than seven years, at least the director level in a large company, or university professor, associate professor To lead the team; the following still needs 1 to 2 Ph.Ds who have graduated for five years, or Managers who graduated from Masters to 78 years; further down, are young doctors and masters. A typical team needs 10-20 people.

Corresponding compensation, the lowest salary of an AI scientist who is employed for 2 to 3 years is 150,000 U.S. dollars a year. This type of person can also get 50,000 to 100,000 U.S. dollars worth of shares in Google and Facebook, which is more than 200,000 U.S. dollars or more. Income has not yet covered medical insurance and other expenses.

The expenses of the overseas technical team are far from affordable for the average company, but the lack of concessions in the country makes Silicon Valley a market to be opened.

Although the chance of success in digging people back from Silicon Valley is very small, domestic startup companies still hold a hope. Alex often ran into the CEO of a domestic AI start-up company at a coffee shop near Stanford. “They basically spend a third of their time a year in Silicon Valley, looking for candidates through various contacts and alumni.”

During the two sessions this year, Baidu founder Li Yanhong, Tencent founder Ma Huateng and HKUST Newsflash chairman Liu Qingfeng, invariably submitted proposals to increase the convenience of introducing overseas talent.

This scene is familiar.

When mobile internet entrepreneurship rose in 2011, the market's hunger for IOS development engineers was no different from that of today's artificial intelligence technicians. However, the technical threshold of artificial intelligence is higher, and the cycle and difficulty of personnel training are even better than the original.

The huge talent demand has spawned a short-term training business. When the mobile internet is hottest, a large number of training companies emerge on the market, claiming that specialist graduates can receive a monthly salary of 10,000 yuan after receiving a training course of a month or two. In that year, there were even many cases of such promises.

Today, the same thing is happening in the artificial intelligence industry.

July is the founder of July Online, an artificial intelligence and data online education. He used to be a blogger who loved to research technical pen questions. In 2012, he noticed that the big companies such as BAT began to examine data structure algorithms in interview questions, and there are more and more people who are interested in algorithms and machine learning.

In 2015 he founded "July Online," and he said that there are now over 20,000 paid students. Half of the trainees are graduate students, and the other half are staff of BAT, Midea, Jingdong, Huawei and small startup companies. "A large Internet company may have tens of thousands of people who are interested in machine learning, but up to hundreds of them are really transparent in the data-related departments." Although July provides basic teaching, not experts, talent Is moving to AI field.

According to Alex, the founder of TalentSeer, an AI recruitment service provider, it was an extremely unbalanced high-price era that lasted no more than three years. In the three years, more AI talents will be cultivated, overheated investment will be corrected, and even many AI companies may disappear after three years (A or B), and some experienced high-quality talents will return to the market again. in.

But in any case, three years is enough for people to boil it once.

Lenovo Chromebook 100E Gen4

Lenovo Chromebook 100E Gen4,Lenovo Chromebook LCD Hinge Cap, Lenovo Chromebook 100E Gen4 Bottom Cover,Lenovo 100e Chromebook Keyboard

S-yuan Electronic Technology Limited , https://www.laptoppalmrest.com