Evan Szu joins Sunny Ray for a wide-ranging conversation about being an early AI pioneer, tracing his path from writing machine learning code in 1995 to founding Gamma Prime today. Evan explains how his father co-founded the International Neural Network Society in 1987 and how neural networks evolved from simple regression models into the technology powering self-driving cars and chatbots. He shares his journey through commodity futures trading, a Stanford PhD, two biotech startups including a breast cancer detection algorithm sold to Ro, and a stint as an actor on Friday Night Lights before becoming CEO of Gamma Prime, which combines AI and blockchain to modernize private market investing. Evan debunks the myth that AI can simply be told to make money trading, discusses the speed of job displacement from coding AI tools, and explains why the biggest AI breakthroughs are happening outside language models, particularly in cellular modeling for drug development. The episode balances technical depth with candid reflection on the human cost and opportunity of rapid AI advancement.
A 1990s AI pioneer turned actor turned biotech founder breaks down what AI can and cannot actually do.
What is Gamma Prime and why should people care?
Most people know public market investments like stocks and bonds, but there's a whole other fragmented universe of private investments including private equity, private credit, hedge funds, pre-IPO and secondaries. Gamma Prime exists to help people make sense of that universe.
You started writing machine learning code in 1995. What did that even look like back then?
AI was a weird sideshow in mathematics. My dad co-founded the International Neural Network Society in 1987 and crammed it into me at dinner. I invented my own degree at Brown combining computer science and chemistry. My first job was at Capital One doing debt valuation with simple regression models, basically the humble roots of everything AI does today.
What pulled you from academia into trading, and how did AI trading actually work out?
It was the other way around, I traded first, starting in 1997 on E-mini futures before going to Capital One. Around 2005 I tried using AI to trade, which is the first thing every AI person attempts. It doesn't work that way because markets change constantly, unlike cats, which always look like cats. You need to constrain the problem to a tiny specific question.
Gamma Prime is your third startup. What's the thread connecting all three?
Machine learning and AI connect all of them. The first two were biotech startups, one doing breast cancer detection with AI that we sold to Ro, and another doing drug development in Singapore that didn't take off. Gamma Prime combines AI with blockchain because private markets are unregulated and perfect for decentralization.
Do you think much about the machine economy?
I worry about it for humans. Imagine training expensive years to be a software coder because you thought computers would always need people, then AI takes over coding jobs within six months. Past job displacement, like electrification, took years to transition. This is happening at a speed we've never seen before, and I don't know what those people will do.
How has the role of ML changed from your first venture to Gamma Prime?
The big shift is language models, which finally bridged AI to humans who didn't need to learn math to access it. But the real innovation isn't in language models, it's in things like self-driving cars and cellular modeling for drug development, which don't involve language at all and could dramatically cut the cost and time of bringing new medications to market.
Building something daring? Sunny talks to founders like this every day. Fifteen minutes to see if your story belongs on the stage.
Claim your pre-interview