Artificial intelligence has become the new gold collar of Silicon Valley
Author:|Brilliant autumn.
Edit | lee shin ma
Title | ic
It's no secret that Microsoft, Google, Nvidia and other tech giants are pushing their AI ambitions up, but running fast doesn't necessarily mean eating meat.
Judging from the fourth quarter quarterly reports just released by various companies, some companies have already enjoyed the dividends brought by artificial intelligence, and some companies are still paving the way for artificial intelligence, even at the expense of making sacrifices.
Previously, the market was overly optimistic about the performance expectations of these giants, but the gap between the results and expectations has brought some headwinds to the stock price movements of some companies. As of January 30, the total market value of AI-related companies had evaporated by 190 billion yuan, or about 1., according to a report in the United States3 trillion RMB.
To date, concepts such as generative AI, big models, etc. are"Smoking guns"Still"Duds"No one can say for sure, but despite this, the field of artificial intelligence has ushered in a moment of involution. If 2023 is just a year of effort, then 2024 will be directly missing those who have done well in the field of artificial intelligence, which may be a step towards the top.
*：ic **
First of all, the financial reporting"Artificial intelligence content"High.
At the end of January, the United States**"Seven tech giants"The report card for the fourth quarter was released, and overall, they did not"Transfers"Investor sentiment, but the performance related to artificial intelligence is still worth a little **.
Microsoft: Using artificial intelligence, Rs 1.7 billion per day.
First, let's look at Microsoft's strategic performance for the second quarter of fiscal 2024 (23Q4 of calendar year): revenue of 6202 billion yuan, a year-on-year increase of 18%; Attributable net profit 2187 billion yuan, an increase of 33% year-on-year, equivalent to a daily profit of 1.7 billion yuan.
The company's three divisions (Personal Computing, Productivity & Business Processes, and Intelligent Cloud) all saw double-digit revenue growth, with the Intelligent Cloud segment in particular reaching 2588 billion yuan, an increase of 20% over the previous year.
At a time when everyone is generally developing, Microsoft's revenue and profit growth can be described as high"Soaring into the sky"Artificial intelligence is undoubtedly the main driver. Microsoft officials said that the 6% growth rate of Microsoft's Azure cloud platform in the quarter was attributed to artificial intelligence, double the 3% growth rate in the first quarter.
However, perhaps because investors had high expectations for the generative AI transformation, Microsoft's after-hours stock price also fell 1% after the earnings report.
Google: Artificial intelligence is developing, but not as expected.
Google's parent company, Alphabet, reported a 13% year-over-year increase in revenue to 863$1 billion. This is Google's fastest revenue growth quarter since the start of 2022, after four consecutive quarters of rapid sales growth. But Google's fourth-quarter cloud business revenue was 919.2 billion yuan, an increase of only 15% year-on-year, a new low in the past ten quarters, Google did not even provide data on the specific contribution of artificial intelligence to the cloud business.
Google lost 108.8 billion yuan (about Rs 773.1 billion) in market value in after-hours trading** by more than 6% in after-hours trading due to lower-than-expected AI development and advertising revenue.
But Google is catching up: on December 6, 2023, Google unveiled its largest and most powerful multimodal macromodel, Gemini"gemini"；On January 18, 2024, Google first showed Samsung its new generative AI feature, Circle to Search. On January 18, 2024, Google unveiled its new generative AI feature, Circle to Search, to Samsung, which the industry considers to be a revolutionary application of Gemini.
Google's CEO intends to focus his greatest efforts on investing in AI and continues to incorporate a variety of emerging generative AI tools into Google's main product line. To that end, Google executives are taking a rigorous inventory of the operations of the business lines to identify areas where they can cut costs.
AMD: Hope for AI chips.
Leading chip company AMD reported fourth-quarter revenue of 61700 million yuan, slightly higher than analysts' expectations of 61$300 million. However, weaker-than-expected guidance for the first quarter of 2023 resulted in the company's shares trading in after-hours trading** by nearly 6%.
However, AMD CEO Lisa Su said at a conference that sales of the recently launched Mi 300x AI processor should reach 3.5 billion yuan, up from the earlier 2 billion yuan. It is worth noting that Wall Street analysts have the most optimistic expectations of $8 billion and the lowest at $4 billion.
Why are the judgments so different? Analysts' ** refers to this year's AI chips, the market size is slowly expanding, and at present, NVIDIA accounts for at least 90% of the share, and the market is generally optimistic that AMD can rely on MI 300X to get a certain share and put pressure on NVIDIA.
Second, downsizing and recruitment go hand in hand.
While several major companies have seen varying degrees of growth in their profits, layoffs have also been the subject of this period.
In the case of Google, according to its official performance report, the company laid off more than 12,000 employees in 2023 for the whole of 2023, and severance and other expenses were about 2 billion yuan.
In January, Google laid off nearly 1,000 more employees, including a core engineering team, a Google Assistant virtual assistant team, and the hardware team that makes Pixel phones, Fitbit watches, and Nest thermostats. Google CEO Pichai said there are more layoffs to come as the company continues to shift its investments to AI:"In order to create the ability to invest in AI, we have to make difficult choices. "
Google isn't alone: Silicon Valley tech giants have been firing companies since late 2022"It doesn't matter"and continues today.
*：ic **
The news that Microsoft's gaming division will lay off more than 1,900 people, or about 8% of the division's 22,000 employees, has come as a shock to the industry. The Activision Blizzard team was hit the hardest by the fiasco, and Blizzard's president and design director has left. Some analysts believe that layoffs are also to save labor costs and better develop important strategic directions such as artificial intelligence.
January's big layoffs also included Amazon laying off hundreds of people at its Prime Video and MGM studios, as well as more than 500 workers on streaming platform Twitch. The owner of Amazon said:"Companies have identified opportunities to cut or stop investing in certain areas, while looking to increase spending on some of the most impactful content and product lines. "
In fact, these giants are not the only tech companies that are laying off workers in the global AI wave. According to the layoff tracking** layoffsAccording to the latest data from the FYI, more than 80 tech companies have announced layoffs so far in 2024, affecting more than 20,000 employees. In 2022 and 2023, global tech companies laid off 160,000 and 260,000 jobs, respectively, according to the data.
The layoffs have been accompanied by more AI-related job openings.
Amazon has prescribed a whopping $34 for a senior manager of applied science and generative AI030,000 yuan high salary; Netflix, a streaming platform, posted a position with a salary of up to 900,000 yuan in August 2023, recruiting people with extensive experience in machine learning platforms; Google plans to further increase the rewards, perhaps in the future with a special one for the company's top AI researchers"Reward pools", there are ** reports that the reward is as high as millions of dollars.
Tech giants"Forced"Make some structural adjustments to adapt to market changes, talents are more in line with development needs, eliminate employees who are not focused on projects, and talents who understand both artificial intelligence and products are becoming more and more popular.
As Vasant Dhar, a professor of information systems at New York University, puts it:"Artificial intelligence is the new Wall Street, and now it's time for big tech companies to make a lot of money; It is a next-generation ATM. It's all about intelligence, and it's all about intelligence in the future. There is an oversupply of really good talent"。
Third, the commercialization of AI is uncertain.
In order to catch the fast train of artificial intelligence, or to run faster, large companies have invested money to increase their spending on artificial intelligence. According to IDC, global companies invested more than 19.4 billion yuan in generative AI in 2023, and it is expected to double to 40.1 billion yuan in 2024, a year-on-year increase of 1067%。
But the reality is that no one has made money from it yet, except for Nvidia and Microsoft.
After betting heavily on the success of OpenAI, Microsoft has successively integrated the capabilities of the latter's great model into its own productivity tools, search engines, operating systems, and cloud services, forming a platform for enterprises, C-end consumers, and developer customers"Full encirclement"。Its Microsoft 365 Copilot, Microsoft's AI tool, has been on the path since its launch: Office 365 enterprise users (E3, E5, Business Standard, and Business Premium) who want to use Copilot will have to pay an additional 30 yuan per month on top of the original price.
While other companies didn't even feel the shadow of the great model, Microsoft's business model emerged and realized. Microsoft's total market capitalization also surpassed Apple's for a time, becoming the world's largest company by market capitalization.
Another one than Microsoft"More ruthless"The player is Nvidia (NVIDIA).
From an industrial perspective, the AI industry based on large models is fueling the explosion of the global AI server market, and the GPUs produced by NVIDIA are supporting AI server computing"Heart"。Crucially, any company that relies on large AI models to make money has to get a piece of the pie with Nvidia.
*:* screenshots.
Nvidia's share price has also benefited from this, with Nvidia's share price reaching a staggering 240% at one point in 2023, a feat that not only made its share price easily outperform the S&P 500 (S&P 500), a benchmark stock index in the United States, but also the top seven tech giants for the year.
However, while some people are working in full swing, others feel like they are in an ice cellar.
In the field of artificial intelligence, the hegemony of giants has accelerated the competition in this field, and the living space will be further compressed for small companies that lack basic technology and financial support to develop their own chips or develop their own artificial intelligence models.
At present, among the most recognized models and gathered participants, whether it is the MaaS model, the open source model, the platform service or the software service, there are large and small development bottlenecks such as slow execution speed and inability to meet user needs. In addition, although some R&D institutions have been relatively mature in large-scale model technology, they are not familiar enough with the landing scenario, and the mining of the true value of the application is still in the process"Early clinical stage"。
To complicate matters further, for the players on the big model track, a lot of money and time need to be invested in the early stage, but the returns are often unpredictable. The most direct example is that due to the difficulty of obtaining high-end chips and the lack of public computing power construction, computing power leasing has entered a new round of price increase cycle, which will lead to a further increase in the R&D cost of large models.
In addition, the safety and ethical issues of large-scale models also put some pressure on commercialization, and without corresponding countermeasures and frameworks, this tool can only stay at the theoretical level or in a laboratory environment, and cannot put its real value into practice.
AI currencies as a whole still have a long way to go, and the specific path to commercialization needs to become the actual technology that companies need to solve problems.
But it is foreseeable that by 2024, technology giants will continue to dominate the development direction of the artificial intelligence track, and launch fierce competition. At the same time, the development of artificial intelligence will pay more attention to the combination of application and landing scenarios"Killer"application, the commercialization model will be further accelerated.