启明创投于WAIC 2025再发AI十大展望：围绕基础模型、AI应用、具身智能等

Core Viewpoint - Qiming Venture Partners is one of the earliest and most diversified investment institutions in the AI field in China, having invested in over 100 AI projects, covering the entire AI industry chain and promoting the rise of several industry benchmark companies [2]. Group 1: AI Trends and Forecasts - Forecast 1: In the next 12-24 months, a context window of 2 million tokens will become standard for top AI models, with more refined and intelligent context engineering driving AI model and application development [3]. - Forecast 2: A general video model is expected to emerge within 12-24 months, capable of handling generation, reasoning, and task understanding in video modalities, revolutionizing video content generation and interaction [4]. - Forecast 3: In the next 12-24 months, the form of AI agents will evolve from "tool assistance" to "task undertaking," with the first true "AI employees" entering enterprises, participating in core processes such as customer service, sales, operations, and R&D [5]. - Forecast 4: Multi-modal agents will become increasingly practical, integrating visual, auditory, and sensor inputs for complex reasoning and task execution, achieving breakthroughs in industries like healthcare, finance, and law [5]. Group 2: AI Infrastructure and Interaction - Forecast 5: In the AI chip sector, more "nationally designated" and "nationally produced" GPUs will begin mass delivery, while new AI cloud chips focusing on 3D DRAM stacking and computational fusion will emerge in the market [6]. - Forecast 6: In the next 12-24 months, token consumption will increase by 1 to 2 orders of magnitude, with cluster inference optimization, terminal inference optimization, and soft-hard collaborative inference optimization becoming core technologies for reducing token costs in AI infrastructure [6]. - Forecast 7: The shift in AI interaction paradigms will accelerate in the next two years, driven by reduced reliance on mobile screens and the rising importance of natural interaction methods like voice, leading to the birth of AI-native super applications [7]. - Forecast 8: The potential for AI applications in vertical scenarios is immense, with more startups leveraging industry insights to quickly achieve product-market fit, adopting a "Go Narrow and Deep" strategy to differentiate from larger companies [7]. Group 3: AI Business Process Outsourcing and Robotics - Forecast 9: The AI BPO (Business Process Outsourcing) model is expected to achieve commercial breakthroughs in the next 12-24 months, transitioning from "delivery tools" to "delivery results," expanding rapidly in standardized industries like finance, customer service, marketing, and e-commerce through a "pay-per-result" model [8]. - Forecast 10: Embodied intelligent robots will first achieve large-scale deployment in scenarios such as picking, transporting, and assembling, accumulating vast amounts of first-person data and tactile operation data, creating a closed-loop flywheel of "model - ontology - scene data" that will drive model capability iteration and ultimately promote the large-scale landing of general-purpose robots [8].