Most companies experimenting with AI have yet to see real financial returns. In a recent Reddit discussion, I noted that while AI tools make individuals more capable, they have not made organizations more profitable. Clients and customers expect faster results for the same price, leaving margins unchanged. The real issue is structural: productivity gains are happening faster than value capture, and soon there will be a struggle over who claims the eventual surplus.

---

A few days ago, I joined a Reddit thread that asked a simple question: “Has your company adopted AI in a way that has provided either cost savings or profit yet”?  The answers were telling. Most people said no. Their companies were experimenting with Copilot or ChatGPT integrations, but no one could point to measurable returns yet. I chimed in with a few thoughts from my own experience running a tech company and product team, and the discussion quickly turned to a deeper question: who actually benefits from AI productivity gains?

One commenter argued that employees who learn to use large language models (LLMs) effectively should renegotiate their pay because they can now do more, faster, and better. It is a fair point. Those who can truly leverage AI are rare, and rarity creates leverage. But it also sparked a more nuanced reflection.

AI has undoubtedly made individuals more capable, but that does not automatically mean their companies are making more money. In my agency, for example, our developers can now complete projects faster with AI assisted tools, but our clients have not suddenly decided to pay us more because of that. Instead, they expect faster delivery for the same price. Likewise, in our product business, no customer has ever said, “This AI feature is so efficient, we would like to pay you extra.” Instead, expectations rise while margins stay flat.

This is the paradox of early AI adoption:

• Productivity gains are real but diffuse.
• Cost savings are visible but not yet realized.
• Value creation is happening faster than value capture.

In other words, the surplus has not arrived yet. And when it does, there will be a struggle over who claims it: employers, employees, or the platforms themselves.

AI is still in its alignment phase inside most organizations. It is making individuals smarter, not yet systems more profitable. Until companies integrate AI structurally, into workflows, data pipelines, and decision models, the productivity it creates will not easily translate into financial returns. But the fight for surplus value has already begun. And as history shows, whoever controls the structure usually wins.

…