The blurring lines between hedge funds and AI labs

The lines between hedge funds and AI labs are blurring as firms compete for elite researchers. Moonshot, a Chinese AI startup, released its Kimi K3 model for public download, highlighting the rapid commoditization of frontier AI. Large platforms have shifted towards building proprietary infrastructure and data, rather than simply using existing models.

Hedge funds are now competing not just with each other for talent, but with AI labs for top mathematicians and computer scientists. Millennium, for instance, is looking to launch its own bespoke AI research lab to expand the development and application of its current technologies. Meanwhile, Morteon Capital Partners is exploiting regional hotspots for AI talent to create its own research lab.

Hedge funds, such as Two Sigma, are developing their own factor analysis platforms like Venn, garnering attention from allocators due to their unique modelling approaches. However, quant funds' advantage may diminish as the middle of their models, including software, risk modelling, and mathematical reasoning, can be executed by large language models (LLMs).

Hedge funds must decide whether to partner with third-party vendors to grow their AI capabilities or develop research models in-house, a strategic dilemma contributing to the growing intertwinement between quantitative investment and AI model development. Developing a sufficient and scalable AI capability comes with significant capital outlay, making it cost-efficient for larger platforms but not feasible for smaller platforms.

Despite the capital large platforms pledge, only a few will be able to create models with a material difference in performance. Funds that demonstrate these capabilities may stand out in the eyes of allocators, who increasingly view auditing a firm's AI integration as a crucial part of the due diligence process.