AI Development Trends
Open-Source Models and Low Costs Drive AI Adoption
Last updated
Open-Source Models and Low Costs Drive AI Adoption
Last updated
In recent years, artificial intelligence (AI) technology has experienced unprecedented growth. Breakthroughs in large models and the rapid expansion of the open-source AI ecosystem have significantly lowered the barriers to AI adoption, enabling its application across a wide range of scenarios.
The rapid adoption of large language models such as ChatGPT and DeepSeek has further strengthened AI’s presence in the market. ChatGPT, developed by OpenAI, has rapidly accumulated a vast user base since its launch. As of February 2025, its daily active users (DAU) in the United States reached 14.9 million. Meanwhile, DeepSeek, a Chinese AI startup, has gained widespread attention with its open-source strategy and cost-effective models. As of February 4, 2025, DeepSeek had amassed 160 million users, experiencing explosive growth. Its DAU exceeded 20 million, and its cumulative downloads reached 16 million within just 18 days—setting an industry record with an almost "one million users per day" growth rate. The model has topped download charts in Apple and Android app stores across 140 global markets, expanding its worldwide influence.
Beyond individual users becoming increasingly familiar with large language model-based AI technologies, DeepSeek’s cost-efficient training and open-source model approach have enabled more startups to thrive in the AI sector, further fueling AI’s rapid advancement. According to Grand View Research, the AI Agents market is projected to grow at a compound annual growth rate (CAGR) of 45.1% from 2024 to 2030.
Despite the rapid advancements of AI in the Web2 space, several fundamental issues persist, including data monopolization, data privacy concerns, centralized computing power, and closed ecosystems. These limitations have concentrated AI development in the hands of a few major tech companies. While DeepSeek’s open-source model partially addresses the problem of AI’s closed ecosystem by reducing technical barriers and promoting decentralization, it does not fully resolve challenges such as computing resource dependency, data silos, lack of ecosystem integration, and difficulties in model iteration and maintenance.