On August 20, 2026, All View Cloud's 2026 Digital AI Ecosystem Conference grandly opened in Lishui, Nanjing. Wang Tingfu, Chairman and Founding Partner of Xingfu Capital, delivered a speech titled , deeply analyzing the development stages of the AI technological revolution, trends in industrial differentiation, and paths for enterprise restructuring from a capital perspective, providing strategic guidance for home appliance and furniture enterprises to embrace AI.

AI Enters Acceleration Phase, Marking the Start of the Fourth Technological Revolution

Wang Tingfu pointed out that AI has entered the agent stage, signaling its acceleration. Over the past year, weekly AI inference token consumption has grown 20-fold, with agent tokens accounting for over 50% of total consumption. By late May, AI agent traffic surpassed human page views (visitor traffic) for the first time. 'AI is the general-purpose technology of the 21st century, representing the fourth technological revolution after the steam engine, electricity, and electronic information.' The development of general-purpose technologies typically progresses through five stages: technological exploration, infrastructure building, application explosion, market bubble, and diffusion. Currently, AI is at the intersection of the early infrastructure phase (2023-2029) and the application explosion phase, with 2026 marking the first year of AI application explosion. AI's development speed far exceeds previous revolutions: while electrification progressed in years and the internet in half-years, AI's cycles are measured in months—30-50 times faster than electrification and 2-5 times faster than the internet.

Intensified K-shaped Differentiation and Industrial Restructuring Under the Solow Paradox

Under the AI wave, industries are exhibiting significant K-shaped differentiation. AI-related sectors continue to explode, with fixed investments in six major U.S. data centers reaching $800 billion by 2025, and China entering a multi-trillion-dollar investment cycle during its 15th Five-Year Plan period. In large models, Anthropic's annualized revenue soared from $9 billion last year to an estimated $100+ billion this year. Meanwhile, non-AI industries face severe pressure: from Q4 2023 to Q1 2024, profits of AI-related companies grew 60.9%, while non-AI companies saw profits decline 23.5%. However, Wang Tingfu cautioned that AI has not yet significantly boosted overall societal productivity, embodying the 'Solow Paradox'—where general-purpose technologies contribute limitedly to total factor productivity during infrastructure phases, with significant improvements only occurring during application explosions. During electrification, Ford's assembly line increased output efficiency 9.6-fold, and the U.S. saw its fastest decade of manufacturing growth (1919-1929). 'General-purpose technologies are necessary but not sufficient for productivity gains; institutional support and process restructuring are also required.'

Three-Step Enterprise Restructuring: From '+AI' to Deep Restructuring

Facing the AI wave, Wang Tingfu proposed a three-step path from simple to deep restructuring. The first step is '+AI' simple restructuring, using AI as a cost-reduction and efficiency-enhancement tool—applying new tools to old businesses. Examples include AI-generated copywriting, report analysis, legal reviews, and programming as external systems to improve process efficiency. Simultaneously, identify internal high-frequency, repetitive, and data-rich scenarios for pilot projects to demonstrate AI's value to all employees.

The second step is deep business process (business process) restructuring, integrating AI into core business processes and adopting new logics for new businesses. First, digitize business processes and build a data mid-platform, then embed models to aid decision-making. His investment in 'Power Steward' improved microgrid power utilization efficiency by 15% through AI-driven intelligent allocation. The third step is product and service restructuring, shifting from feature delivery to outcome and service delivery, redefining product boundaries and business models. Examples include selling temperature-control services instead of air conditioners, or offering leasing + AI predictive maintenance for forklifts. Ekom Smart Driving transformed from 'selling autonomous mining trucks' to 'contracting mine transportation services,' capturing 55% market share—surpassing all competitors combined. Additionally, Wang emphasized the importance of organizational restructuring.

AI-native organizations are highly flattening (flat), such as OpenAI's structure of hundreds of 6-person 'pod' teams, where 20+ people developed Sora. Traditional enterprises must promote AI participation in task allocation, establish a 'human + AI + token' joint evaluation mechanism, and spin off mature AI businesses for independent operation. 'Facing this once-in-a-century technological revolution, it's better to define new rules with AI than be disrupted by it,' Wang concluded. The greatest AI value lies not with tool providers but with industry reshapers. Traditional enterprises, equipped with real industrial scenarios, deep process expertise, and long-term industry knowledge, are the core force for unlocking AI's massive potential.

Original article by All View Cloud. Unauthorized reproduction or scraping of this content for AI large model training or other purposes is strictly prohibited.