According to news on October 23, on the occasion of the 10th anniversary of the artificial intelligence company SenseTime, Xu Li, chairman and CEO of SenseTime, recently issued an internal letter to all employees, mentioning for the first time the company's newly established "large device-large model- "Application" trinity strategy , and announced that it will build a more centralized and efficient organizational structure around strategy and core resources, and accelerate the process of making organization and management lighter.
According to a report from Data Intelligence Frontline today, after the letter to all employees was issued, SenseTime started organizational adjustments and layoffs that day . At present, layoffs are still in progress, "Shanghai has vacated large conference rooms," and the scale of layoffs is unknown.
According to reports, SenseTime has focused on retaining two businesses related to Ririxin's large models and large devices, and the remaining business lines, including security, autonomous driving, and medical care, have begun to adjust . “Some organizational structures have been basically cut down, including entire departments.” A SenseTime employee said that she saw a huge adjustment.
The report also mentioned that the compensation given this time was N+1 . "Last year there were still N+2.5." An employee said. Because this layoff was "one size fits all, with no breathing space", it caused dissatisfaction among employees.
Inquiring into the financial report of SenseTime Technology, we learned that in the first half of 2024, SenseTime achieved revenue of 1.74 billion yuan, a year-on-year increase of 21%; the net loss decreased to 2.457 billion yuan from 3.123 billion yuan in the same period last year.
Among them, generative AI has become SenseTime’s main source of revenue . In the first half of 2024, SenseTime's generative AI business reached 1.051 billion yuan, a year-on-year increase of 255.7%, and contributed more than 60.4% to total revenue.
Attached is a letter from all members on Shang Tang’s 10th anniversary:
"SenseTime starts again on its 10th anniversary: focus, unity of knowledge and action"
Dear Shangtang classmates:
Hello everyone! SenseTime celebrates its tenth birthday this month. At the same time, the artificial intelligence community also ushered in a major milestone: the Nobel Prize in Physics recognized work "using physics to advance artificial intelligence", while the Nobel Prize in Chemistry recognized "using artificial intelligence to advance protein structure prediction" achievements. This marks the interaction between artificial intelligence and science, each serving as a research purpose and a research tool, indicating that a new paradigm in a series of subject areas is about to begin. Under the portrait of Alan Turing on the new fifty-pound banknote is a sentence: "This is only a foretaste of what is to come, and only the shadow of what is going to be.” – This is perhaps the best comment on our time.
We believe and look forward to the arrival of the era of general artificial intelligence. Pragmatically speaking, we have taken two steps: what is often called traditional AI 1.0 and generative large model AI 2.0. Generally, AI 1.0 is regarded as specialized intelligence, focusing on single tasks and information processing; while AI 2.0 is regarded as general intelligence, emphasizing multi-tasking and content generation. However, although this description is simple and easy to understand, it is not completely accurate because there is no clear boundary between general and special applications, and the application of AI must be scenario-based after all, such as the application of large generative models in vertical fields. In our view, an important difference between AI 1.0 and AI 2.0 is the change in AI cost structure.
In the traditional AI 1.0 era, the main cost of model production lies in the investment of R&D personnel. In the era of generative large model AI 2.0, the cost of model production mainly lies in the investment of computing resources . As the scaling law is verified in large language models, multi-modal models, video generation models, and slow thinking reasoning processes, the cost of producing and using large models can be directly equivalent to the consumption of computing resources. In short, the popularization and commercialization of generative large model AI requires efforts to reduce the production and use costs of large models. This requires combining large models to iterate and optimize computing power. It also needs to be based on the characteristics of computing resources. Iterative large model design and application.
Therefore, in the field of generative large model AI, SenseTime's core strategy is to achieve seamless integration of computing power devices (SCO), large models and applications (CNI), to use applications to drive models, and to use models to drive the optimization of computing power. We have established a trinity strategy of "large devices - large models - applications", aiming to improve the efficiency of computing resource use and serve our customers well through order-of-magnitude optimization. We are committed to becoming the large model service provider that best understands computing power, and the computing power service provider that best understands large models.
In the traditional AI field, we will make full use of our core capabilities of visual perception and multi-modal models, pool resources, clarify directions, and serve both domestic and international markets through a set of R&D investments.
At the organizational level, focusing on strategy and core resources, we will build a more centralized and efficient organizational structure , promote the centralized and intensive investment of resources, and accelerate the rejuvenation process of organization and management. As SenseTime celebrates its tenth anniversary and the era of general artificial intelligence is approaching, let us regain our original intention to start a business together.
Thanks to every SenseTime employee for their dedication and efforts over the past ten years. I hope that we will continue to maintain the tenacity, courage and optimism of the past, proactively embrace changes, unswervingly explore the path of "technological originality and closed-loop industrial value", and work hard in the next ten years to jointly define and create general artificial intelligence era!