Business Breakdowns · Colossus | Investing & Business Podcasts

How Investors are using AI

·49 min·3 clips

AI can map first- and second-order exposure from tariffs to supply chains and competitors.

1. Business Breakdowns examines how investors are using AI, with David Plon of Portrait Analytics as the guest. 2. David Plon is the founder of Portrait Analytics, and the host says his investing background makes him useful for describing investor pain points. 3. The episode asks how AI can improve research workflows without replacing the parts that build conviction. 4. David says his career ran through Barclays on the trading floor, Slate Path Capital, and Balpost before Stanford Business School. 5. He says deep learning felt inevitable to him at Stanford, before the transformer era, and he eventually left to build Portrait full time. 6. He frames the main research bottlenecks as idea generation, initial context building, and ongoing monitoring. 7. He says generalists often know there are dozens of ideas in their sweet spot but cannot tell which ones they are working on at a given time. 8. He says AI helps with the early triage step, when an idea should be killed before deeper work begins. 9. He says AI can also move deeper analyses earlier in the pipeline, including CEO compensation reviews and multi-year guidance checks. 10. He gives an example of laying out three or four years of management guidance to test credibility and incentive alignment. 11. He says management may technically beat guidance while still revising down expectations quarter after quarter. 12. He says AI is useful for portfolio monitoring because investors can pull in sparse but relevant signals from customers, suppliers, and competitors. 13. He uses Expedia to show how a travel investor may care about hotel pricing, OTA distribution, and consumer demand rather than every Marriott release. 14. He says AI can help with idea sourcing around external developments like the tariffs announced in April of last year. 15. He describes a harder sourcing task: finding companies that fit a nuanced mental model such as a formerly strong business hit by a temporary hiccup. 16. He says translating a gestalt investing instinct into a query is difficult, but it can identify a 10-out-of-10 idea profile from prior history. 17. The conversation shifts into prompt writing, and David says an effective prompt should read like an email to a smart person overseas with limited context. 18. He says good prompts usually include task, background, output format, guidelines, and domain knowledge, especially skepticism toward management commentary. 19. Investors who like practical AI workflows, investment screening, and research process design would find this episode useful. 20. Investors who want product demos, or who dislike process-heavy workflow talk, may skip it.

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Business Breakdowns