Enterprising Investor Brian Pisaneschi, CFA: Demystifying AI for Smarter Investing
14 snips
Aug 18, 2025 Brian Pisaneschi, CFA, a Senior Investment Data Scientist at CFA Institute and recognized as an Asset Management Game Changer, dives into the future of AI in investing. He discusses retrieval-augmented generation (RAG) and how it improves AI accuracy while combating misinformation. Pisaneschi highlights the limits of AI in complex quantitative tasks, the blend of automation with traditional methods, and the irreplaceable value of human creativity in finance. His insights are essential for anyone looking to navigate the evolving landscape of AI-driven investments.
AI Snips
Chapters
Transcript
Episode notes
How RAG Grounds LLMs
- Retrieval-augmented generation (RAG) grounds LLM outputs by pulling relevant text chunks into the model's context window.
- RAG reduces hallucinations and supplies updated real-time information beyond a model's training cutoff.
LLMs Are Not Calculators
- Large language models are autoregressive text generators not native calculators or table parsers.
- They often hallucinate on complex numeric extraction and struggle with varied table structures.
Use Agentic Workflows
- Break complex tasks into smaller steps and connect LLMs to dedicated tools for each step.
- Use APIs, calculators, or Python interpreters so the model calls accurate functions instead of guessing.