Thinking Machines published a manifesto titled *The Future Worth Building Is Human*. At first glance, it reads like another vision statement in an increasingly crowded AI landscape. But beneath the discussion of models, alignment, and customization lies a much bigger idea — one that extends far beyond AI itself.
The manifesto argues that intelligence alone isn't enough. The knowledge that drives businesses, industries, and societies doesn't live inside a centralized model. It lives inside people. It is created every day by individuals solving problems, making decisions, learning from experience, and adapting to situations that no dataset can fully capture.
Reading it, I wasn't thinking about AI. I was thinking about a decade spent building AI for commerce. **If there's one industry that has always operated on distributed intelligence, it's retail.**
For decades, brands have invested billions trying to understand customers. They built CRM platforms, loyalty programs, recommendation engines, customer data platforms, and sophisticated analytics systems. The assumption was simple: if we collect enough data, we'll understand what customers want.
Yet anyone who has worked on a sales floor knows that's only part of the story.
The most valuable knowledge in commerce has never lived inside a dashboard.
It lives in the beauty advisor who immediately recognizes a customer's concerns before they say a word. It lives in the stylist who instinctively knows when to recommend a different silhouette. It lives in the dermatologist who adjusts a skincare routine based on subtle observations that never make it into structured data.
These decisions aren't random. They're built on years of experience, thousands of conversations, and an intuition that's difficult to explain, let alone encode into a spreadsheet.
This is the kind of knowledge economists have long described as *tacit knowledge* — expertise acquired through experience rather than instruction. It is local, contextual, and constantly evolving. More importantly, it is distributed. No single organization can centralize it completely because it exists across thousands of individuals interacting with customers every day.
Technology Tried to Eliminate This Complexity
For years, technology has tried to eliminate this complexity.
Digital commerce optimized for efficiency. Self-service became the goal. Chatbots replaced conversations. Automation became synonymous with progress.
The assumption was that software scales better than people. In many ways, it does. But something unexpected happened.
As digital experiences became increasingly frictionless, consumers found themselves overwhelmed rather than empowered. Every retailer offered recommendations. Every website promised personalization. Every platform optimized for conversion. Intelligence became widely available.
What became scarce was confidence.
The more information consumers received, the more valuable trusted guidance became.
The Next Chapter Won't Be Defined by the Smartest Model
This is why I believe the next chapter of AI won't be defined by who builds the smartest model.
It will be defined by who best amplifies human expertise.
That may sound counterintuitive at a time when headlines celebrate fully autonomous AI agents capable of replacing entire workflows. But commerce has always operated differently than pure automation.
Buying decisions are rarely based solely on information. They are based on trust.
People don't choose a dermatologist because they have access to more data than anyone else. They choose them because they trust their judgment. The same is true for beauty advisors, financial advisors, architects, trainers, and countless other professionals whose expertise is built through relationships rather than algorithms.
Artificial intelligence doesn't eliminate the need for trust — it increases its value. As AI makes information nearly free, trust becomes the scarce resource. And scarcity creates value.
The Next Competitive Moat
This shift also changes what organizations should consider their greatest competitive advantage.
For years, companies believed their most valuable assets were their products, their technology, or increasingly, their data.
I think the next competitive moat will be something far more difficult to replicate.
Collective expertise.
Every customer conversation. Every recommendation. Every objection overcome. Every successful outcome.
These interactions are more than just transactions. They are moments where organizational knowledge grows. Until recently, that knowledge disappeared as soon as the interaction ended. AI now gives companies the opportunity to preserve it, learn from it, and continuously improve — without removing the human who created it in the first place.
That's a fundamentally different vision from replacing people with machines.
It's about making people exponentially more effective.
From Digitizing Workflows to Capturing Judgment
The companies that understand this won't simply deploy AI assistants. They'll build systems where human expertise and artificial intelligence continuously improve one another. Their competitive advantage won't come from owning the largest model, but from cultivating the richest network of human knowledge.
The last decade was about digitizing workflows.
The next decade will be about capturing judgment.
Not replacing it. Not standardizing it.
But scaling it — while preserving the uniqueness that makes every organization different.
Distributed Intelligence Wins
History suggests intelligence becomes more powerful when it's distributed.
Markets thrive because knowledge is distributed. Innovation flourishes because ideas emerge from diverse perspectives rather than central planning. The internet became transformative because publishing was decentralized. Open-source software accelerated innovation by allowing thousands of contributors to improve systems together.
AI may follow the same path.
Instead of one intelligence serving everyone, we may see millions of specialized intelligences shaped by millions of experts, organizations, and communities.
If that future unfolds, the winners won't necessarily be those with the biggest models or the most compute.
They'll be the organizations that know how to combine three things better than anyone else:
Human expertise. Artificial intelligence. Trusted relationships.
Commerce Has Always Been the Transfer of Confidence
Commerce has never simply been about selling products.
At its core, commerce has always been the transfer of confidence from one person to another. AI won't change that. It will simply allow trust — and the human expertise behind it — to scale in ways that were never possible before.
Looking back, I think we have been asking the wrong question about AI in commerce. The question was never whether machines could replace experts. It was whether they could make every expert exponentially more valuable.
That's the future I've become convinced we're building at Replika Software — one where AI doesn't compete with human expertise, but compounds it.
And for companies that learn to scale trust instead of replacing it, that may become the most durable competitive advantage of all.
