Interview with Reflexivity CEO Jan Szilagyi on how AI is being used by hedge funds, especially in global macro, to synthesize data, test ideas faster, reduce blind spots, and improve execution. He argues AI is already creating productivity gains, but finance alpha will not disappear immediately because market relationships are unstable, data is sparse in many macro areas, and human judgment still matters.
Watch on YouTube ›Get the market thesis, key claims, assets, contradictions, and follow-up questions from any financial video — then unlock a version personalized to your portfolio, watchlist, and favorite speakers.
This is a Forward Guidance interview with Jan Szilagyi, CEO and co-founder of Reflexivity, an AI software firm serving hedge funds. Szilagyi explains his career path from math at Yale to working alongside Stanley Druckenmiller, taking a PhD in quant finance at Harvard, and later managing global macro portfolios at Fortress with Mike Novogratz. He says Reflexivity was built around the idea of reflexivity itself: markets and fundamentals influence each other, and AI can help investors and decision-makers understand that feedback loop faster and with broader context. A major theme is that large language models are useful for finance because they can perform multi-dimensional synthesis across many data points quickly, but general-purpose chatbots are not reliable enough for high-stakes analysis. …
Near term, the actionable setup is around specialized AI tools improving research speed and trade discovery, especially in macro and commodities. The risk is confusing generic chatbot output with robust, auditable analysis.
Over the next few months, adoption should broaden among funds that can integrate proprietary data and workflow controls, which may widen the gap between early adopters and everyone else. The key validation signal is repeated success on sparse-data questions without relying on fragile prompts or manual cleanup.
Structurally, AI looks likely to become a permanent layer in investment decision-making, but not a full replacement for human portfolio construction anytime soon. The deeper regime shift is a productivity and capex boom that may be disinflationary over time even as it creates real-world bottlenecks in compute, energy, and commodities.
Reflexivity is designed to help investors understand market and fundamental feedback loops in real time.
The guest explains the name and says the system is meant to give feedback to decision-makers about how actions alter outcomes.
Large language models help finance because they can synthesize many market possibilities at once rather than linearly.
He contrasts human step-by-step analysis with LLM multi-dimensional synthesis.
AI can improve macro analysis by expanding small sample sizes through analogs across similar countries and crises.
He says Reflexivity can define similar experiences across EM countries and create larger datasets from comparable cases.
What was the value of doing a PhD in quant finance after working with Stanley Druckenmiller?
He says Druckenmiller did not think the PhD would be value-additive for a hedge fund career, and he agreed it was not especially useful at the time. In hindsight, though, it helped him systematize what he had been doing and shaped how Reflexivity came together.
What is Reflexivity, and why did you choose that name?
He says the name was intentional and tied to the Soros-style reflexivity idea: markets reflect fundamentals, but fundamentals also respond to markets. The product is meant to give analysts and portfolio managers real-time feedback so decision-makers can adjust course.
Why did you leave global macro investing to start Reflexivity?
He saw, even before ChatGPT, that AI could unlock the value trapped in financial data. His core thesis was that large language models could synthesize many market implications at once, much faster than a human analyst can.
Unlock the full claims, asset map, scores, related transcripts, follow-up questions, and AI chat — shaped around your portfolio, watchlist, favorite speakers, and risks.
