Chinese Quant Firm Goku Expands Into AI for Scientific Discovery
Shanghai-based quantitative trading firm Goku Technologies, managing assets worth over 15 billion yuan (US$2.1 billion), is moving beyond finance by launching an AI-focused subsidiary, Shanghai AllMind Artificial Intelligence Technology. CEO Wang Xiao states the company’s goal is to apply AI to address scientific challenges rather than compete in the consumer AI market.
Since 2017, Goku has invested in AI research, concentrating on model training and integrating AI into trading and scientific applications. The infrastructure originally built for financial data analysis now supports research in materials science, chemistry, and biology. Notably, Goku collaborated with Shanghai Jiao Tong University to submit a research paper to the prestigious Conference on Neural Information Processing Systems (NeurIPS), introducing a new AI training method that reportedly outperforms existing techniques.
Quantitative Funds’ Evolution into AI Research
The transition from quantitative trading to AI research follows a pattern seen throughout financial history. The finance industry was among the first to explore AI, dating back to the 1950s with automation efforts in accounting and banking. The 1980s saw heavy investments in expert systems aimed at market prediction, though early efforts fell short of expectations.
Today’s quantitative funds like Goku benefit from advanced data infrastructure and algorithmic expertise developed for trading, providing a strong foundation for AI research. This shift reflects broader trends in quantitative investing, where machine learning expands analytical capabilities beyond traditional statistical methods.
China’s Strategic AI Push and Finance Sector Expertise
Goku’s AI initiative is part of China’s broader strategy to advance AI across sectors, backed by a government commitment reportedly totaling $8.2 billion. Quantitative funds expanding into AI research are a growing trend in China’s AI ecosystem. For example, High-Flyer Quant launched DeepSeek in 2023, followed by Goku’s AllMind in 2025.
These finance-rooted AI ventures emphasize rigorous mathematical methods and benefit from significant capital, enabling them to compete with established tech companies in fundamental AI research. Goku’s submission to NeurIPS marks the first time a Chinese quantitative fund has contributed to this leading AI conference, signaling a serious commitment to global scientific AI development.
AI-Driven Scientific Discovery as the Next Frontier
Goku’s focus on scientific applications highlights a shift in AI’s role—from business optimization to accelerating research and discovery. Autonomous AI systems are beginning to conduct scientific research independently, as shown by a Japanese AI that authored a peer-reviewed paper without human input.
This trend is gaining traction across research institutions, with projects like CENSAI dedicated to using AI to transform scientific methods and speed up breakthroughs. While AI-related scientific papers tend to receive higher citations, adoption varies across disciplines. A key challenge remains: many scientists lack sufficient training in AI techniques, which could limit the full potential of these tools.
- Quant funds bring strong data and algorithm backgrounds that support AI research.
- China leverages these finance-backed AI efforts as part of its national AI strategy.
- AI's impact on scientific discovery is growing but uneven across fields, with education gaps to address.
For researchers interested in expanding their AI skills, exploring targeted AI training courses can be valuable. Platforms like Complete AI Training offer resources tailored for scientific and research professionals.
Your membership also unlocks:
