The River Paradox And Why Your Approach to Technology Might Be Dangerously Wrong

What if the very thing everyone believes will save their business is actually destroying it? It’s a provocative question, but one that demands our attention in an age of relentless technological churn.

Picture this: you're standing on the bank of a wide, fast-flowing river. This river represents the ceaseless current of technology – a decade ago it was deep learning, then reinforcement learning, then transformers, and now the great flood of Large Language Models (LLMs). With each new current, you cautiously dip a toe in the water. You run a pilot project, a proof of concept. You are sensible, you are prudent. You are testing the temperature before you commit to swimming.

But here’s the uncomfortable truth that Ritavan, bestselling author of Data Impact, discovered after a decade in the data trenches: whilst you’re busy dipping your toes, your competitors aren’t just learning to swim across the river – they’re building bridges. This isn’t another tired story about ‘digital transformation’ or ‘AI adoption’. This is about why the conventional wisdom around technology might be fundamentally flawed, and why the businesses that survive the coming decade will be those brave enough to challenge everything they think they know.

Ritavan’s journey to this startling conclusion was anything but conventional. He began his career not as a data scientist or a consultant, but as a trader. In the unforgiving crucible of the financial markets, he learned a brutal lesson: if everyone can do what you’re doing, it doesn’t matter how well you do it. True advantage is proprietary; it is unique. This foundational, first-principles thinking about competitive advantage would later revolutionise how he approached corporate strategy.

After a decade operating in the data field, watching company after company chase the latest shiny object, he felt a growing unease. “I had to take a step back from operating for half a year to get mental clarity,” he reveals on the Project Flux podcast. It’s a telling admission. In the frantic, hair-on-fire reality of quarterly targets and endless delivery, there is often no time for altitude. Yet, it is only from altitude that you can truly see the battlefield.

That period of reflection culminated in his book, Data Impact, and its SLASSOG framework (Save, Leverage, Align, Simplify, Optimize, Grow). More profoundly, it led to a series of contrarian insights that cut against the grain of almost everything we hear about technology. Today, Ritavan is obsessed with what he calls “the leverage part” of his framework – the art of compounding small advantages over time to achieve monumental, long-term wins. “Everyone’s chasing quick wins,” he explains, “but you cannot string a whole bunch of quick wins together and get a big win in the long game. It’s like trying to win a marathon by sprinting 100 metres, over and over again.”

This brings us to the technology adoption paradox, a concept best illustrated by a story about one of the world’s most famous investors. An executive at Berkshire Hathaway once breathlessly reported to Warren Buffett that a new type of textile mill had been invented that was twice as efficient as their own. The executive expected excitement, perhaps a major new investment. Buffett’s reaction was the polar opposite. After verifying the claim, his decision was simple: exit the business entirely.

As his partner Charlie Munger would later explain, Buffett understood a fundamental market dynamic that most of us miss. If you can buy a piece of technology off the shelf, so can everyone else. The vendor who sells it to you will happily sell it to your competitor tomorrow. The result? The overall supply in the market increases, but the demand for your product or service remains the same. Basic economics dictates what happens next: prices are driven down, and margins are crushed.

In this cycle, there are only two clear winners. The first is the technology vendor, who, as Ritavan puts it, is “skimming off the most predictable part of your profit margin” through subscriptions and licensing fees. The second is the rare company that understands this game and refuses to play, instead focusing on building a proprietary advantage that cannot be bought. Today’s AI boom is following this script to the letter. “I’m willing to bet in a few years it’ll be something else,” Ritavan observes. “We’re constantly walking along the river and dipping our toes in the water, but never swimming across it.” This should be a sobering thought for any professional services firm whose primary value proposition is now built on leveraging the same generative AI tools that are available to everyone.

This critique of conventional wisdom extends to one of the most pervasive clichés in modern business: the idea that “data is the new oil.” It sounds smart, but Ritavan argues it’s a dangerously misleading analogy. “If you’re saying data is a commodity, like oil, then how can it at the same time be the driver of unique value creation for you?” he asks. “You can’t have it both ways.”

Companies that truly win with data don’t treat it as a commodity to be hoarded. They see it as a proprietary asset, a vital component in a larger, integrated system designed to win in their specific market. They think in terms of moats, competitive advantages, and proprietary edges – the language of investors, not the language of industrial-age resource extraction.

He illustrates this with a powerful example from the construction industry. Imagine you are managing inventory for a large project. The traditional approach is to create a point forecast: you’ll need exactly seven units of a particular material on a specific day. The finance department loves this; it’s clean, simple, and easy to budget. But it’s also, as Ritavan notes, “empirically invalid.” It assumes a level of certainty that simply doesn’t exist in the real world.

A more rigorous approach is to think probabilistically. There might be an 80% chance you need seven units, but a 15% chance you need nine, and a 5% chance you need fewer. This isn’t about making things more complex for the sake of it. It’s about enabling smarter decisions. If you calculate that being just two units short will delay your entire project by a month, the cost-benefit analysis changes dramatically. Suddenly, the “optimal” decision might be to keep fourteen units in stock, even though seven is the most “accurate” forecast. The goal isn’t forecast accuracy; it’s decision optimisation. You’ve moved from simply predicting the future to building a system that is robust and resilient in the face of an uncertain future.

This brings us to the elephant in the room: the AI bubble. With some estimates suggesting that a third of recent US GDP growth has been driven by capital expenditure into AI infrastructure, and with market indicators like the Buffett Indicator at historic highs, the parallels to the dot-com boom are hard to ignore. Ritavan offers a nuanced take. Hype, he explains, is a powerful mechanism for coordinating decentralised actors to make massive capital investments – buying up Nvidia chips, building vast data centres, and training ever-larger models. It has happened with railways, with electricity, and now with AI.

But these hype cycles have a predictable pattern. If the promised productivity gains don’t materialise – if companies continue to merely dip their toes rather than fundamentally re-engineering their businesses to swim – a painful correction could be inevitable. The opportunity lies with those who can look past the hype and use this moment to build genuine, lasting advantage.

We are at an inflection point. The choice is to continue the cautious, toe-dipping approach, perpetually chasing the next wave of technology, or to learn to swim across the river whilst others are still testing the temperature. This conversation with Ritavan barely scratches the surface of his SLASSOG framework and the profound implications of his thinking.

The full podcast episode reveals layers of insight that space constraints prevent us from exploring here. You'll discover why Ritavan believes most data maturity models are fundamentally flawed, how he applies military doctrine thinking to business strategy (including why taking "cavalry to an armoured brigade fight" perfectly captures most companies' approach to AI), and the specific mathematical frameworks he uses to model trade-offs in high-stakes decisions. He shares candid stories about companies that have successfully implemented his approach, as well as cautionary tales of those who ignored these principles at their peril.

Perhaps most importantly, you'll hear his detailed breakdown of each component of the SLASSOG framework – not just the theory, but the practical steps for implementation. How do you actually "leverage" data for long-term competitive advantage? What does it mean to "align" your data strategy with your business model in a way that creates genuine moats? These aren't abstract concepts; they're battle-tested approaches that have helped organisations move from quarterly firefighting to decade-long competitive dominance.

If you're tired of chasing quick wins and ready to build something that compounds over years, not quarters, this conversation will fundamentally alter how you think about data, technology, and competitive advantage. Listen to the full episode to discover why the companies that win the next decade will be those brave enough to swim across the river whilst everyone else remains standing on the bank, forever testing the water temperature.