The AI Investment Mirage

You know how the AI narrative goes. It is always “move fast and build smart,” but at the end of the quarter, only a few are capturing value.

The Entrepreneur piece steeped this in research from MIT’s “The GenAI Divide: State of AI in Business 2025.” It was based on 150 interviews with AI leaders, an examination of 300 AI apps, and a survey of 350 employees. The finding is arresting: U.S. enterprises have spent between $35 billion and $40 billion on AI, but 95 percent of them report no measurable impact or return. Only 5 percent are extracting value, sometimes jumping revenue from zero to $20 million in a year, mostly by focusing narrowly on one use case and using third party tools effectively.

Those that win, won by starting small and picking the right fight. Those that fail, failed because AI simply did not slot into daily workflows. Tools like ChatGPT stall when they do not adapt to existing processes or integrate with enterprise systems. AI did not conform, so the business ignored it.

So what does this mean for engineering leaders and CTOs today

First, the dot com style hype is already fraying. Market watchers are tightening their jaws. Wall Street took notice when this MIT report reverberated through the tech sector, triggering sell offs and bubble talk. The same research also shows that companies buying AI tools perform better than those building them in house.

Second, this is not just a metric problem, it is a design problem. AI is often deployed as a novelty project or vanity sprint. Without context or the ability to learn over time, it yields zero ROI. That reinforces what economists have termed the modern productivity paradox: massive investment, little business impact. It echoes Robert Solow’s observation about computing in the 1980s, when he famously said, “You can see the computer age everywhere but in the productivity statistics.”

Hiring great developers used to be the bottleneck. Now it's about finding people who can move fast, collaborate across tools, and co-build with AI.

TECLA helps U.S. tech teams hire senior-level, English-proficient engineers from across Latin America at up to 60% less cost, and in your time zone.

Want to see how fast you could scale?

What is the path forward for leaders who still believe in AI

There is one, but it requires humility and discipline.

Pick one friction point you understand deeply, maybe invoice reconciliation, backlog categorization, or customer triage, and focus the AI on solving that. Use battle tested tools rather than building your own unless your team has iron clad use cases. Build for adaptation. Tools must learn how your business actually operates instead of forcing new ways of working.

Track real ROI, ideally impact on profit or employee time savings, not just adoption. If your pilot will not move the needle now, pause it before it becomes a vanity project buried in the budget.

Bend your leadership mindset away from scale and toward focus. Winners have one use case, clear integration, proper metrics, and durable ownership models. Others wander in pilot purgatory.

The real revolution is not rushing to the next AI frontier. It is building AI that lives in the business, not next to it.

More to come…

✔️ Is the AI boom finally starting to slow down? (The Guardian)

✔️ American companies have invested billions in AI initiatives, but have basically nothing to show for it (The Radar Pro)

✔️ From AI Pilots to Enterprise Value: A CDO’s Playbook (The CDO Times)