How I Use AI for Investment Research

Key Concepts

• AI-Assisted Research: Using Large Language Models (LLMs) for preliminary data gathering and sector analysis.
• Cross-Verification (Triangulation): The practice of querying multiple AI models simultaneously to identify inconsistencies or hallucinations.
• Hallucination: A phenomenon where AI models generate false or fabricated information.
• Investment Workflow: The integration of AI into the initial stages of financial decision-making rather than automated execution.

AI Integration in Investment Research

The speaker discusses the evolving role of Artificial Intelligence in their investment research process, noting a significant improvement in model utility over the past six months. While previously considered "useless," AI tools have become a staple for preliminary information gathering.

1. Methodology: Multi-Model Cross-Verification

The speaker employs a comparative methodology to mitigate the risks of AI inaccuracies. By inputting the exact same prompts into multiple platforms—specifically Grok, ChatGPT, Claude, and Gemini—the speaker can:

• Identify Hallucinations: Detect when a model is "pulling information out of nowhere."
• Validate Data: Cross-check responses across different architectures to ensure consistency and reliability.

2. Scope and Limitations

The speaker maintains a clear boundary regarding the depth of AI involvement in their financial activities:

• Current Application: AI is used primarily for the "initial researching stage." This includes learning about specific stocks, sectors, or industries to establish a foundational understanding before conducting deeper, manual analysis.
• Excluded Applications: The speaker explicitly avoids using AI for:
  • Crafting proprietary investment or trading strategies.
  • Automated, 24/7 algorithmic trading (bot trading).
• Frequency: The tools are utilized approximately once per week to facilitate learning and research.

3. Strategic Perspective

The speaker views AI as a "starting point" rather than a decision-making authority. The core argument is that AI serves as a research assistant that helps the user understand where to look next, rather than a tool that dictates the final investment decision. The speaker emphasizes a cautious, manual-heavy approach, ensuring that AI is only a preliminary step in a broader, human-led investment framework.

Synthesis and Conclusion

The speaker’s approach to AI in finance is characterized by pragmatic skepticism. By leveraging multiple LLMs to cross-verify data, the speaker effectively manages the risk of AI hallucinations. The primary takeaway is that while AI is highly effective for rapid information synthesis and sector education, it is not yet trusted for autonomous strategy development or execution. The speaker’s workflow highlights a transition from viewing AI as a novelty to utilizing it as a functional, albeit limited, research tool that requires human oversight at every stage.