Implied CEO on the Limits and Capabilities of AI for Investing
In this episode, Brett Caughran and Khe Hy sit down with Ying Hua, founder and CEO of Implied and a former PM at Citadel and Balyasny, to get into where AI actually fits inside an investment process and where it still falls short. Ying makes the buyside case for why the horizontal foundation labs won't "eat" finance, and why domain knowledge is the piece they won't build.
We get into:
The case for vertical AI tools over the foundation labs in finance
Why public-market investing is one of the hardest domains for AI to learn
What "institutional-grade" data really takes and why Implied transcribes earnings calls itself
Whether a non-coding analyst can actually build tools with Claude Code
Where dashboards help, where they break, and what replaces the static dashboard
The line between AI synthesis and human judgment, and what stays human
How AI is reshaping Excel and the daily investment workflow
What the most advanced users are automating right now
We're not coming at this as "experts" with all the answers. We're in it every day, testing, breaking things, and trying to understand where this is going. The goal of the podcast is simple: bring you along as we learn, and give you a clearer view of how AI is actually being used in investing. If you work in equity research, at a hedge fund, or on the buyside and you're trying to make sense of AI, this is a good place to start.
Chapters (Timestamps)
00:00 – Intro: Ying Hua and the Vertical AI Bet
00:55 – Ying's Background: Goldman, Citadel, Balyasny → Implied
01:52 – Why Starting an AI Investing Company in 2023 Was Contrarian
02:20 – The Two Catalysts: GPT-4 and the Gap No Tool Filled
03:45 – Horizontal Foundation Labs vs Vertical Finance Agents
04:12 – Why Stock-Picking Is One of the Hardest Domains for AI
08:30 – Is Pattern-Matching Its Own Kind of Knowledge?
08:51 – "We Had a Great Quarter": The Training-Data Problem
11:07 – The Institutional-Grade Data Challenge
11:53 – Why Implied Runs Its Own Live Transcription
16:02 – A Scraper for Car Crashes: Finding Data in the Cracks
19:15 – Is Data Scraping an AI Skill or Just Easier Code?
21:05 – Can a Non-Coder Build It in Claude Code?
25:21 – Dashboards, Agent Debt, and the Tokenomics Problem
29:19 – The Hard Part: Customizing AI to Domain Knowledge
32:32 – Synthesis vs Judgment: What Stays Human
37:00 – The Materiality Problem and the SpaceX Example
39:24 – Fixing Excel: The Investor's Workbench Meets AI
45:32 – What the Power Users Are Actually Doing
46:07 – The Next Leg: From Chatbot to a Junior That Learns
49:46 – Outro: From Search to Agentic Process Cloning
Want to actually build these workflows yourself? The AI Accelerator is Fundamental Edge's 6-month cohort for investors who want repeatable AI workflow.
Learn More below:
https://www.fundamentedge.com/ai-accelerator
Follow Invest with AI on Youtube: https://www.youtube.com/@FundamentEdge