ChinAI #345: A Three-way Race for China's AI Super-App

2022-03-07 · Source: ChinAI Newsletter · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure, Data Science & Analytics · Depth: Intermediate, quick

Summary

ByteDance's Doubao has emerged as China's leading AI super-app, surpassing 100 million daily active users and processing 63 trillion tokens daily, a 200% increase in six months. This success follows an initial lag in ByteDance's LLM development compared to Baidu, Huawei, and Alibaba. Doubao's breakthrough leveraged ByteDance's Douyin platform and a unique voice call function, which involved extensive dialect collection and stylized dialogue training. Tencent, initially cautious, is now aggressively catching up with its Yuanbao chatbot, based on the Hunyuan model, and has integrated DeepSeek. Alibaba positions itself with full-stack AI capabilities, including its PPU Zhenwu 810E chip, which secured its first external major customer in 2025. Chinese tech giants hold significant advantages over startups in talent acquisition, compute resources (over 100,000 GPUs each), and proprietary internal data, though they still face challenges in cutting-edge research compared to global leaders like OpenAI.

Key takeaway

For product managers evaluating AI super-app strategies in competitive markets, ByteDance's success with Doubao highlights the critical role of integrating AI products with existing high-traffic platforms and developing unique, localized features like advanced voice capabilities. Your team should focus on leveraging proprietary data and platform synergies to create distinct user experiences, rather than solely pursuing benchmark scores, to achieve sustainable user growth and market leadership.

Key insights

ByteDance's Doubao leads China's AI super-app race, driven by platform integration and unique voice features.

Principles

• Platform integration drives AI product adoption.
• Data contamination skews model evaluation.
• Full-stack capabilities offer competitive advantage.

Method

ByteDance's Doubao team collected dialects down to district/county levels and conducted stylized dialogue training to give its AI assistant a distinct personality, enhancing user engagement.

In practice

• Integrate AI features with existing high-traffic platforms.
• Prioritize real-world performance over benchmark rankings.
• Standardize data cleaning for model training.

Topics

• Chinese AI Competition
• AI Super-Apps
• Large Language Models
• AI Talent Acquisition
• AI Compute Infrastructure

Best for: Product Manager, Investor, CTO, AI Product Manager, Director of AI/ML, Business Analyst

Related on AIssential

• 😺ChatGPT is about to look completely different — OpenAI is consolidating its AI products into a "superapp" to dominate enterprise workflows and counter fragmentation.
• Is This the Best AI Video Model in the World? — ByteDance's Seed Dance 2.0 sets a new benchmark in AI video generation with integrated audio and advanced capabilities.
• 645: How China's AI Punches Above Its Capex, Google's Brain Drain, Neural Software, SpaceX's $6.3B Compute Deal, A24, Five Gens of TPUs, Amazon's Robots Learn English, and Three Beams of Light — Fast-follower strategies enable rapid AI advancement with lower capex by exploiting pioneer discoveries and stable hardware architectures.
• Can open-source beat OpenAI? — Open-source AI models foster global collaboration and efficiency, challenging closed-source dominance through accessible innovation and diverse monetization.
• ChinAI #358: Around the Horn (25th episode) — China's AI ecosystem is dynamic, marked by rapid startup valuations, evolving business models, and significant talent and regulatory challenges.
• Codex, Cursor and the Great Beyond — Intense competition and strategic acquisitions are rapidly reshaping the AI market, challenging early leaders.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by ChinAI Newsletter.