The recent shift in technology, particularly towards generative Artificial Intelligence (AI), has sparked significant discussions in the corporate world, particularly in ChinaSensetime, a prominent player in the previous wave of AI advancements, is making a bold transition into this new arena, marking a critical pivot in its strategy.

On December 3rd, the Chairman and CEO of Sensetime, Xu Li, communicated a transformative vision through a letter addressing all employeesHe introduced a revamped organizational structure encapsulated in a "1+X" modelIn this framework, the "1" symbolizes the core business focus, where Sensetime aims to establish itself as a leader in AI cloud services, a domain where large-scale infrastructures, foundational models, and AI applications work seamlessly togetherThis demonstrates their ambition to advance further in the Computer Vision (CV) sector by developing versatile visual models applicable across various scenarios.

The "X" represents a diversified ecosystem of companies that are being restructured for greater innovation and specialization

This matrix will include ventures such as “Mojing” for smart vehicles, “Yuanluobo” revolving around household robotics, along with initiatives in smart healthcare and intelligent retailingIn this new model, each entity within the ecosystem will appoint its own CEO, responsible for driving its unique business objectives, thereby promoting agility and accountability within its operations.

This strategic reshuffle stems from a reevaluation of the company’s direction amid rapidly evolving technological landscapesThe restructuring seeks to position Sensetime to tackle opportunities and challenges presented by what is increasingly being termed the AI 2.0 era—characterized prominently by generative AI capabilities.

Previously, during the last transformative wave in AI, Sensetime distinguished itself through its prowess in facial recognition and visual algorithms, earning its place among the “Four Little Dragons of AI” in China

• Alibaba Sells HighLuck Retail for HKD 13.138 Billion
• Offline Retail: A Once-in-Two-Decade Opportunity?
• The Biggest Risks in Financial Markets for 2025
• Dollar Rises Despite Fed Rate Cut: Here's Why
• Lincoln in China: Merger Fears Resurface

Their successful IPO on the Hong Kong Stock Exchange in 2021 further solidified its status as a tech giantBy the end of 2022, as the generative AI movement embodied by tools like ChatGPT gained momentum, Sensetime began amplifying its investment in large AI models and applications.

The company has unveiled advancements in its “Riri X” model, leveraging computational power derived from over 10,000 GPUs since the end of 2022. By April 2023, they rolled out the 1.0 version of Riri X, marking it as one of the first large language models released by SensetimeThis was followed by an impressive upgrade to version 5.5 in July 2024, which claimed comparable capabilities to GPT-4o, becoming China's first multimodal real-time interactive large model.

This latest strategic overhaul further solidifies Sensetime's commitment to embedding generative AI within its organizational framework, reflecting a shift towards a core specialism in this space.

Celebrating its tenth anniversary in October, Xu Li expressed optimism about the unfolding era of general artificial intelligence, stating that the company has already made significant strides towards developing traditional AI 1.0 and transitioning towards generative large model AI 2.0.

One pivotal distinction Xu emphasized between these two eras of AI development pertains to changes in cost structures

In the traditional AI 1.0 era, expenses incurred largely stemmed from research and development personnelIn contrast, the generative AI landscape of 2.0 has shifted this focus predominantly towards computational resources and infrastructure utilization.

In pursuit of generative large model AI's widespread adoption and commercial viability, it is imperative to lower the costs associated with production and deploymentThis initiative demands an iterative approach that fine-tunes both model design and practical application alongside the evolution of computational resources.

As such, Sensetime’s strategic thrust in the domain of generative large model AI centers around achieving a seamless integration of significant computational apparatus (SCO), large model frameworks, and application efficacy (CNI). The goal is to let application demands drive model development, while optimizing computational resources in response to these evolving demands

To facilitate this, Sensetime has realigned its organizational structure focusing on these strategic necessities.

Xu Li underscored three critical dimensions of the new structure's strategic significance: Firstly, it emphasizes generative AI as the core of Sensetime's business model, promoting quicker pathways to profitability and cash flow stabilitySecondly, leveraging group infrastructure offers synergistic advantages that empower ecosystem entities to capture vertical market opportunities more effectivelyLastly, the independent CEO structure aligns with market-driven growth dynamics.

Confronting the challenges posed by what many identify as the "Fourth Industrial Revolution" surrounding generative AI, Sensetime brings to bear years of technical expertise, robust AI infrastructure, and a well-established application ecosystemAccording to financial reports, revenues from Sensetime’s generative AI segment reached an impressive 1.051 billion yuan in the first half of 2024, showcasing a staggering year-on-year increase of 255.7%. This segment has now emerged as a pivotal pillar in the company’s portfolio, comprising 60.4% of its overall revenue.

Despite these promising developments, the road ahead is fraught with challenges for Sensetime