Giants are laying off employees and shifting manpower to AI that doesn't need much manpower

The five major technology giants in the United States, Microsoft, Apple, Alphabet, Amazon, and Meta, will release their financial reports for the first time in 2024. Investors and the public are starting to care about how much money they have thrown at AI, how much money they have made, and if they continue to throw money, whether they will continue to lay off employees.

The U.S. economy is still strong, and judging by the stock price in the last year, Silicon Valley seems to be enjoying the boom brought by generative AI, but why are tech giants and big factories constantly laying off employees, round after round?

Silicon Valley faces a stark new reality: layoffs are the very reason these giants are recognized by the capital markets.

Cloud makes money, cloud "launders money"

The big model makes the cloud business profitable first. Microsoft's intelligent cloud business achieved revenue of $25.9 billion in the most recent quarter, up 20% year-over-year, and Azure and other cloud revenue increased by 30%, with AI contributing 6 percentage points. Google Cloud grew 26% to $9.2 billion and posted its first full-year profit. Amazon's AWS division reached $24.2 billion in sales, up 13% year-over-year, and it is expected that AWS's full-stack generative AI services will bring tens of billions of dollars in revenue in the future, which is equivalent to rebuilding AWS.

According to Synergy Research, the large model and generative AI boom pushed the revenue of the entire cloud infrastructure market to $74 billion in the fourth quarter of 2023, up $12 billion from the same period last year and $5.6 billion from the third quarter, the largest quarter-over-quarter growth experienced by the cloud market. The cloud infrastructure market grew from $212 billion in 2022 to $270 billion for the full year.

The growth of the three cloud businesses all shows a certain effect of what Silicon Valley calls "cloud money laundering", that is, huge investments in large-scale model enterprises, most of which are converted into cloud computing quotas. In this way, Microsoft's tens of billions of dollars in OpenAI, Amazon and Google's investment in Anthropic, and about $6 billion, a large part of them have become cloud computing businesses. The growth of the intelligent cloud business has further pushed up the stock price. Cloud computing technology companies invest in large model companies, which has become a routine and its ecological strategy. This alarmed the U.S. Federal Trade Commission (FTC), which last week gave them 45 days to provide information on agreements related to recent cooperation and investment.

Big models are changing the landscape of the cloud market. Synergy has calculated the market share of IaaS, PaaS and managed private cloud services, and Microsoft is significantly increasing its market presence. The application of large models has just begun, so Google and Amazon will still get enough room for growth.

Trends in the market share of cloud services

Note: Only IaaS, PaaS, and managed private cloud services are included.

Large models, the tools that the giants first invented for themselves

The big model seems to have made the giant's own business profitable, but it is still in its infancy. Whether it's SaaS giant Microsoft, search giant Google, social giant (advertising giant) Meta, retail giant Amazon, and consumer electronics giant Apple, they all talked about the positives of AI.

Microsoft is using GPT-4 for its new product, Microsoft 365 Copilot, at breakneck speed, for a monthly subscription fee of $30 (GPT-4 is $20), but no user data is disclosed. Copilot enables users to search, write, and summarize tasks up to 29% faster.

Google attributed its steady growth in advertising to Google's investment in generative AI, and said the search generation experience (SEG) was still in its early days, but it had already reduced response times by 40%.

Meta's ad impressions increased 21% year-over-year, and average prices increased 2% year-over-year. Zuckerberg announced that he will integrate AGI into all of his social media apps in the future, and their monthly active users have grown to 3.98 billion.

While acknowledging that strong revenue came from a record-breaking holiday shopping season, Amazon is most pleased with continued business innovation and customer experience improvements, including generative AI's personalized buyer recommendation tools and seller advertising tools, and has high hopes for the newly released generative AI shopping assistant Rufus. The largest cloud computing company, which is not as aggressive as Microsoft and Google in terms of large models, was eventually tempted to invest $4 billion in Anthropic, most of which was in cloud computing.

Apple's main revenue still comes from consumer electronics, but that doesn't mean there is no artificial intelligence, Apple just doesn't talk about it. Apple's M-series chips are the first to have a neural network computing and processing unit, and have supported a large number of AI applications and feature enhancements on the device side, which are integrated into the user experience. Apple's active device installations now exceed 2.2 billion units, and perhaps locally deployed end-to-end models, including mature product lines, the new Vision Pro, and future smart cars, are the way for more consumer electronics users to truly share in the progress of AI.

Apple CEO Tim Cook faced analysts' questions, saying that Vision Pro's technology from chips to displays to spatial computing is driven by AI, and also revealed that "there are some things that we are extremely excited about, and we will talk about them later this year." iOS 18 is expected to be the "biggest" update in the history of Apple's operating system, and in terms of integrating AI features, it may be on par with Microsoft's Windows update, which was released around the same time.

In 2024, expand and throw money

All the giants know that their businesses are increasingly relying on deep learning technology. The competition has just begun, and the R&D spending and capital investment of tech giants on large models will continue to increase. The capital expenditure is mainly to build computing infrastructure, and the R&D expenditure is mainly to poach people and build teams, including large model teams and self-developed chip teams.

2024 is the year of Google's Gemini, and Google wants to use this large model to surpass GPT-4, and its Gemini-Ultra version should be released soon, first for search. Google's capital expenditure, which is mainly used for artificial intelligence, increased by $2.96 billion in the most recent quarter, and its investment in AI in 2024 will also be significantly larger than in 2023.

Meta sees a good chance of catching up. Zuckerberg claims that the company sits on hundreds of billions of publicly shared images and tens of billions of public videos, much larger than Common Crawl's dataset. Its self-developed chip "Artemis" (Artemis) can be deployed in data centers this year for inference, and Meta is expected to save hundreds of millions of dollars in energy costs and billions of dollars in procurement costs every year. Its training chip is still being developed in-house, and it will have to be snapped up from Nvidia before that. By the end of this year, Meta will deploy 350,000 NVIDIA H100 GPUs. The company expects full-year 2024 capital expenditures to be between $30 billion and $37 billion, up to $9 billion higher than full-year 2023.

So far, Nvidia has been the biggest beneficiary of all of this. The big five tech giants are snapping up its latest GPUs, the H100 and H200. I've been busy for a year, and I've been working for NVIDIA. In January, its market capitalization increased by an unprecedented $296.5 billion in a single month to about $1.52 trillion, surpassing the record of $248.2 billion set in May 2023. Nvidia will release its latest quarterly earnings report on February 21.

"Will redirect manpower to places where it is no longer needed"

Despite the huge profits, these giants are structurally shrinking their fronts. They even save money on layoffs, paying billions of dollars in break-up fees to free up more resources to roll up large models. And the more advanced the AI used in their business, the fewer people they will need.

Microsoft CFO Amy Hood has made it clear to investors that it will continue to shift its workforce to AI-first businesses that no longer need much staff.

In 2023, Google laid off more than 12,000 employees and spent $2.1 billion on severance and other expenses. Entering 2024, more than 1,000 positions were cut in a month, and another $700 million in employee severance payments were spent. Meta spent more, reaching $3.5 billion. Of that amount, $2.5 billion came from "facilities consolidation", i.e., the closure and consolidation of offices, and another $1 billion for "severance and other personnel costs", i.e., the dismissal of employees. The company currently has 22% fewer employees than last year. This year, Amazon has also continued to lay off employees and halted some investments, with the recent contraction mainly concentrated in the entertainment sector.

When the tech giant first reported its earnings in 2024, both its CFO and CEO stressed to investors that they were improving efficiency through layoffs. And layoffs have indeed become an important reason for the tech giants' rapid rebound from the 2022 slump. The most typical ones are Meta and Alphabet. Of course, the most ruthless layoffs in 2023 are the layoffs of three-quarters of the workforce after Musk's acquisition of Twitter.

In the capital market, layoffs are not negative news for the company, but glorious layoffs, which show that the company has a clear strategy, is well managed, and can quickly allocate resources to the most effective business and future development prospects.

The capital market is convinced of this. In 2022, the Nasdaq, dominated by tech giants, lost a third of its market capitalization, and in 2023, it rose by 43%, and in January 2024, it rose by another 3%.

Layoffs are the ladder of giant AI progress

Layoffs in Silicon Valley, especially those laid off by large companies, were able to find jobs in the tech world, but now they can only go outside the tech world.

Tech companies laid off 260,000 people last year because they overhired during the pandemic and high interest rates prevented it from investing in tech startups at high costs. Now it's time to get back to normal. However, when layoffs continue in 2023 and extend into 2024, the economic situation is good and interest rates are returning to normal, which makes it feel like the employment base of the tech industry is shaking.

The U.S. added 353,000 jobs in January, double what economists expected. But Google, Amazon, Microsoft, salesforce, ebay, PayPal, and discord are all laying off employees on a large scale.

Among the hundreds of thousands of high-tech personnel who have left science and technology companies, the competition for jobs is becoming more and more intense. And the more senior it is, the harder it is to find a job. For example, software engineering and data scientists from large factories used to wear halos, but now they find a suitable position, often thousands of people apply, and they can always find a new job within 3 in the past, and now many people have not found a new job since they left their jobs at the beginning of last year until this year. This is something that many people haven't experienced in Silicon Valley for 20 years.

The overall economic situation of the U.S. economy is improving, which has also attracted many people who are beginning to say goodbye to your circle completely and seek jobs that can achieve a better work-life balance. The enviable perks of big companies are also being eliminated one after another, such as free laundry, free massages, free food and fitness facilities.

The Washington Post found that "some of the recent layoffs have targeted middle managers who manage teams that have been hit by previous waves of layoffs." Some of them are trying to get back to writing code instead of mentoring others because they think they might be safer. Employees who try to jump ship from one company to another every three or four years to maximize the stock options they can accumulate are now left to lie where they are. But now it's not as safe to write code, and generative AI is getting better and better.

From Nadella to Huang, they are saying that the more advanced AI is, the faster companies will improve efficiency and growth, and the greater their contribution to the economy and the more jobs will be created. But the reality in the field of AI is that a new breed of start-ups is innovating with fewer and fewer people.

For example, Google is laying off a large number of employees on the one hand, and on the other hand, it is increasing the salary incentive plan of millions of yuan for more elite AI researchers. Salesforce is laying off employees on a large scale, but the CEO himself has stepped in to poach people from OpneAI with high salaries. Enterprises are spending billions and tens of billions of dollars on GPUs, expanding data centers, and competing for larger computing clusters.

AI is at the forefront of innovation, and it seems to bring greater computing power and innovation from a smaller number of elites, but the results of this innovation may make the stock prices of large manufacturers soar and the country feel honored, but whether it can ultimately benefit the economy and society remains to be verified.

Where will AI eventually create jobs? If it eventually becomes a general-purpose technology and is applied to all walks of life, real jobs will be created in large numbers in the process of application.

2024 is critical.