Tech

The Fierce Battle for AI Talent in Silicon Valley: Tales of Poaching and Salary Wars

The battle for AI talent in Silicon Valley is going crazy.

Musk himself personally stamped it: “This is the craziest talent war I have ever seen.”

How crazy can it be?

On the one hand, big names like Zuckerberg and Sergey Brin were revealed to be personally involved in the competition for AI talents.

According to The Information, Xiao Zha has already started headhunting: personally writing letters to DeepMind researchers, inviting them to change jobs and join Meta.

The news circulating from Brin’s side is that in order to retain employees who are about to leave and switch to OpenAI, the Google founder personally made a phone call, talking about salary increases and benefits. The main point was that as long as the employees stayed, the conditions would be Everybody can talk.

This is not the end.

On the other hand, Musk’s xAI has reached the point of “stealing people” from its big brother Tesla…

In Musk’s own words, if xAI does not provide an offer, people will be poached by OpenAI.

OpenAI has been offering huge salaries to poach Tesla. Unfortunately they have had some success.

The Ethan Knight mentioned by Musk has not even reached the “supervisor” level. His title at Tesla is machine learning scientist.

The CEO of a big factory personally goes out to grab people

In fact, mutual poaching and talent flow among major Silicon Valley companies are accelerating.

In the last week alone, Meta has lost three senior AI employees:

Including AIGC senior director Devi Parikh, computer scientist Erik Meijer (referred to as Uncle Mei) and Meta AI research scientist Abhishek Das.

The first two have worked at Meta for 7 or 8 years, while Das has been cooperating with Meta during his Ph.D. and joined the technology giant as soon as he graduated.

It doesn’t matter if he leaves, Uncle Mei still left this paragraph on his social media, which really caused thousands of waves with one stone:

If you want to build something cool on top of a big model, there’s really no advantage to being inside a big company.

In addition, according to a report by the Wall Street Journal in June last year, under the impact of the new wave of generative AI, at least one-third of Meta’s AI researchers left due to burnout and lack of confidence in the company.

Under such pressure, it’s no wonder that Zuckerberg couldn’t help but take action personally and use his personal email address to recruit talents for Meta.

Meta’s talent preferential policies don’t stop there:

Now, Meta also provides AI talents with the opportunity to join without interviews.

Some “hidden rules” have also begun to be broken. For example, in the past, Meta would not use promotions and salary increases to retain employees who wanted to change jobs, but now, the situation has changed.

Xiao Zha’s efforts have indeed yielded some results.

While talents are flowing out, new blood from competitors is also joining.

For example, Michal Valko, a senior engineer at DeepMind, announced on X that he would join Meta as the chief engineer of Llama.

Xiao Zha is not the only one who steals people.

As for Google, in addition to Brin, who personally called to retain employees as mentioned above, Pichai, Jeff Dean and others also contributed.

For example, Logan Kilpatrick, the former head of developer relations at OpenAI who recently abandoned OpenAI and joined Google, was invited by Google executives, including the two of them.

Mustafa Suleyman, former CEO of Inflection AI and co-founder of DeepMind, was poached by Nadella.

Crazy poaching with seven-figure annual salary

In this increasingly chaotic trend of grabbing people, salary is of course an important factor.

Even Meta, which has been ridiculed as a “depression”, according to the salary data website Levels.fyi, the median salary of machine learning and AI engineers is close to 400,000 US dollars, which is almost 2.89 million yuan.

After all, in terms of competitors, OpenAI, which Musk named as “spending money”, has a median total salary of US$925,000, including stocks, which is approximately 6.69 million yuan.

Data from Levels also shows that OpenAI pays some senior engineers up to US$1.4 million (approximately RMB 10.13 million).

The Information even got the news: When OpenAI poached people from Google, it promised that the annual salary (mainly in the form of stocks) would be between US$5 million and US$10 million.

Of course, Meta is not unresponsive to this kind of money being thrown at people: Meta also offers annual salaries of US$1 to 2 million to senior researchers hired from outside.

Not only Meta, Musk also couldn’t sit still and said he would raise the salary of Tesla’s AI team.

However, salary is obviously not the only criterion for these top AI talents to choose where to live.

Computing power has also become one of the core attractions.

Perplexity founder Aravind Srinivas revealed that he originally wanted to poach someone from Meta, but because there were not as many GPUs as Meta, the other party declined.

Previously, Zuckerberg revealed that Meta will have more than 350,000 Nvidia H100s by the end of this year.

In the final analysis, it is still resources, resources, resources.

The winner of the talent flow among major manufacturers is yet to be determined. At present, they are still secretly competing with each other in a back-and-forth situation.

Startups have indeed been affected by the first wave of negative impacts:

Inflection AI and Stability AI, two of the most vulnerable companies whose CEOs have left, are facing severe challenges in terms of talent and talent.

The core team of Stable Diffusion collectively resigned from Stability AI at the end of March.

But as Uncle Mei, who left Meta, said, large models are the new infrastructure. The “creativity” in the minds of AI talents is not just a kind of resource.

error: Content is protected !!