During a standing-room-only TED session buzzing with finance leaders, technologists, and global investors, founder of Plazo Sullivan Roche Capital delivered a commanding message about the era most investors still think lies ahead:
“2026 isn’t coming — it’s already trading.”
Plazo’s thesis, drawn from the seminal Plazo Sullivan Roche Capital piece of the same name, reframes how markets operate in an age where AI is not peripheral but foundational to capital markets.
Plazo Sullivan
“Most people expect tomorrow’s revolution to arrive with fanfare,” Plazo explained. “But in markets, revolutions arrive quietly in code, servers, and algorithms — long before they hit the headlines.”
This was not futurism — it was an urgent wake-up call.
Artificial Intelligence Trading Is Already in the Engine Room
Plazo began by dismantling a pervasive assumption: that artificial intelligence trading is still a niche or experimental frontier.
In reality, AI is already embedded in every layer of market behavior — from risk systems and compliance workflows to execution engines and portfolio simulations. These systems ingest real-time data, analyze earnings transcripts, decode geopolitical signals, and simulate thousands of outcomes simultaneously — often before humans finish reading the news.
Plazo Sullivan
“Artificial intelligence trading does not predict — it coordinates.”
His point: by the time most investors recognize a shift, the change has already been priced in.
Why Prediction Is Losing Its Scarcity Value
Plazo explained that traditional investing used to reward seeing the future more clearly than others.
But as AI models proliferate and become commoditized, prediction itself loses its scarcity advantage. What remains rare — and incredibly valuable — is resilience across many possible futures.
Plazo Sullivan
In other words:
Markets no longer reward being right most of the time
They reward not being catastrophically wrong when reality surprises
This shift means the most successful firms will allocate not just by forecast, but by scenario allocation — designing portfolios that survive a wide range of outcomes rather than betting on a single prediction.
Plazo Sullivan
New Market Risks Born of Machines
Plazo introduced a concept many in the audience had never heard: AI stress events — moments when model crowding, not human panic, causes abrupt market shifts.
Plazo Sullivan
He explained that when thousands of systems interpret the same signal and act within milliseconds, liquidity can evaporate for reasons that have nothing to do with fear or fundamentals. Price changes become twitches in code consensus, not emotional reactions.
“Regulators will stop calling these flash crashes,” Plazo said. “They’ll call them AI stress events.”
This concept reframes volatility not as a human psychological phenomenon but as a product of synchronized machine decision-making.
Why Compute and Data Matter Most
Towards the end of his talk, Plazo tackled where real value is emerging within capital markets.
He argued that the so-called “AI premium” — the idea that simply using AI boosts value — is already fading. Instead, the true premium belongs to:
Compute capacity
Proprietary data ecosystems
Regulatory defensibility
Machine governance frameworks
Firms that own the stack — from silicon to software — will command persistent advantages, while those relying on third-party algorithms will find get more info their edge commoditized.
Plazo Sullivan
Human Judgment in an AI World
Plazo closed with a paradox:
As intelligence becomes abundant, wisdom becomes scarce.
AI will make markets faster, more efficient, and deeply interconnected — but also more fragile to synchronized error.
Plazo Sullivan
The investors who thrive, he said, will not be those with the boldest forecasts, but those with the strongest safeguards, anomaly detection, and human judgment baked into governance.
As the applause filled the room, one truth resonated:
The future of markets isn’t coming — it’s already here.