Artificial intelligence dominates the conversation in 2025, but what if AI, as we know it, is missing the bigger picture? That was the central question posed by Nadab Akhtar (Niddy), founder and managing partner of Excite, during his panel at The Lively Grit Daily House at SXSW 2025. Akhtar, whose firm specializes in probabilistic AI-driven trading using quantum mechanics, challenged the conventional wisdom behind predictive AI models, arguing that most current systems rely on outdated assumptions.
His premise? The world is chaotic, and tomorrow doesn’t look like yesterday, so why are so many predictive models still designed as if it does?
The Problem with Traditional Predictive Models
Whether in finance, business, medicine, or logistics, predictive models — AI-powered or manual — operate under the assumption that past patterns can reliably forecast the future. Akhtar compared this to driving at 80 miles per hour while only looking in the rearview mirror. Traditional trading strategies rely on historical data, statistical correlations, or news-driven models to make financial decisions. But in a rapidly changing world, that’s not enough.
He broke down three core trading methodologies most firms use:
- Latency-Based Trading: Competing to be the fastest, reducing execution times by microseconds.
- Headline Trading: AI-driven strategies reacting to news stories.
- Correlation-Based Analysis: Creating statistical models from past market behavior.
While these strategies have fueled high-frequency trading for years, they all share a fundamental flaw: they assume the market behaves in a fixed, predictable way.
Probabilistic AI: A New Way to Think About Markets
Excite’s approach rejects these static models. Instead, it models volatility itself. Rather than reacting to old data or headlines, its probabilistic AI focuses on real-time inputs — like a driver adjusting to traffic in real time rather than relying on yesterday’s road conditions.
Using principles of quantum mechanics, Excite’s AI calculates thousands of potential market scenarios in microseconds, assessing risk and probability in real-time. Akhtar explained that this mirrors how human brains operate: we don’t make decisions based solely on past experience, we instinctively assess risk in the moment.
Quantum Without the Hype
For many, the word “quantum” brings to mind images of futuristic computers running on mysterious principles of physics. But Akhtar emphasized that Excite’s quantum approach doesn’t rely on yet-to-be-developed quantum hardware. Instead, his firm applies quantum-inspired algorithms to financial modeling.
Among these are:
- Data Hamiltonians: Real-time models of market conditions, similar to how physics models energy over time.
- Chattering Algorithms: Like an ultra-smart GPS that constantly recalibrates based on shifting conditions.
- Data Tomography: A method akin to a CT scan, layering real-time snapshots of market movement to build a 3D model of financial activity.
These quantum-inspired techniques allow Excite’s AI to constantly adjust trading strategies based on the market’s “speed and acceleration,” rather than outdated static models.
Why Isn’t Everyone Talking About This?
At the close of his panel, Akhtar tackled the elephant in the room: why isn’t quantum-driven AI a bigger part of the AI conversation? The answer, he explained, is twofold.
First, AI is trendy, and quantum sounds complicated. Many assume quantum is still decades away, believing it requires specialized hardware. Excite’s approach proves otherwise.
Second, developing quantum-inspired AI is hard. While major players like NVIDIA and other tech giants focus on traditional machine learning, developing entirely new probabilistic AI models requires a fundamental shift in thinking — and that takes time.
Looking Through the Windshield
If today’s AI-driven trading firms are looking in the rearview mirror, Excite is focused on the windshield. Akhtar’s quantum-inspired AI models are designed to adapt in real-time, offering a glimpse into what the next generation of financial technology will look like.
As AI continues to evolve, companies that embrace real-time probabilistic modeling may gain a serious advantage over those relying on traditional predictive systems. And while the financial world is the testing ground for this approach today, its applications — from healthcare to logistics — could be just as transformative.
At SXSW 2025, Akhtar didn’t just talk about AI’s future. He made it clear: that future is already here, and it’s moving faster than we think.
Related Articles
Spencer Hulse is the Editorial Director at Grit Daily. He is responsible for overseeing other editors and writers, day-to-day operations, and covering breaking news.
