Under the AI Wave, How Fund Managers Use AI for Investment Research
- The article is a conversation between NDV founder Jason and Jason Kam, founder of Folius Ventures, about how fund managers use AI in research.
- Kam describes a multi-agent AI workflow in which several AI roles debate a topic and a separate research director role synthesizes the result.
- He says AI is useful for scanning public sources, mapping evidence across materials, and handling repetitive tracking work.
- Kam also built BidClub, a searchable library of investment podcasts; the article says the official site listed 29 shows and 2,251 episodes as of August 16, 2026.
- The discussion argues that AI first reproduces much of a fund manager’s workload, but it does not automatically improve judgment or solve the underlying investment question.
- The article presents AI as powerful for research efficiency, while noting its limits in identifying the right problem to ask.
This article is a recorded conversation between NDV founder Jason and Jason Kam, the founder of Folius Ventures. The discussion focuses on how two fund managers who are not programmers have started using AI in their investment research workflows, and what AI can and cannot do for investors.
Kam says the value of AI is not only in automation, but in organizing research work more systematically. He describes using multiple AI roles to examine a question, debate with one another, and then hand the output to a final synthesis role. The article also notes that he uses AI to scan public social media, forums, research materials, and selected podcasts for emerging themes.
The conversation highlights three practical uses: continuous information scanning, building a research map that separates evidence from secondhand claims, and tracking recurring updates and arguments over time. Kam also points to BidClub, a searchable library of investment podcasts, as an example of how AI can make spoken research material easier to search and revisit.
The broader point is that AI is reducing the cost of gathering and processing information, but it does not make judgment effortless. The article argues that AI is first replicating fund managers’ workload rather than replacing their decision-making, and that a key limit is that it can dig deeply into a question without knowing whether the question itself is the right one. From a market perspective, the piece is neutral: it describes a productivity trend in investment research, but does not report a specific trading catalyst or event.
在人工智能浪潮下，基金经理如何用 AI 做投资研究
这篇文章是一场对谈，参与者是 NDV 创始人 Jason 和 Folius Ventures 创始人 Jason Kam。内容围绕两位并非程序员出身的基金经理，如何把 AI 引入投资研究流程，以及 AI 对投资人究竟能做什么、不能做什么展开。
Kam 认为，AI 的价值不只是自动化，而是把研究工作组织得更系统。他介绍了一种多智能体工作流：多个 AI 角色分别研究同一问题、相互辩论，最后再由一个汇总角色整合结论。文中还提到，他会用 AI 持续扫描社交媒体、论坛、研究材料和精选播客，寻找正在形成的议题。
文章归纳了 AI 的三类实际用途：持续筛选信息、搭建研究地图以区分证据与二手转述，以及跟踪公司动态和反复出现的论点。Kam 还提到自己创建了可检索的投资播客库 BidClub，用来把原本需要逐集收听的内容先检索出来再回看原文。
文章的核心判断是，AI 正在降低信息获取和处理成本，但并不会自动提升判断力。它首先复制的是基金经理的工作量，而不是直接替代决策本身；同时它能深入回答问题，却未必知道问题是否问对了。就市场影响而言，这篇报道偏中性，主要反映投资研究效率提升的趋势，没有给出具体交易催化事件。