Everyone is celebrating AI as the great accelerator.

After 30 years inside large corporations, I can tell you what nobody is saying out loud: AI did not solve our oldest problem. It exposed it.

For decades, we called it “business requirements.” Today we call it “the prompt.” The name changed. The discipline did not.
A bad requirement used to produce a useless system in six months. A bad prompt now produces a useless answer in six seconds. Same failure. Faster.

The data is uncomfortable. MIT’s 2025 NANDA report, “The GenAI Divide,” found that despite $30 to $40 billion invested in generative AI, 95% of enterprises see no measurable return.

And the reason is not the technology. MIT calls it the “learning gap”: organizations did not know how to use the tools or design workflows that capture real value.

Translation: the business did not know what to ask. So the technology did not know what to deliver. So nothing landed.
Same problem we had 30 years ago, dressed in new clothes.

If the business cannot articulate the outcome, no model on earth will rescue the project. We did not upgrade the thinking. We upgraded the noise.

Real thinking does not come from a tool. It comes from experience. Experience is knowing which question is worth asking. It is the scar tissue that tells you what to do first and what to leave for later.

You cannot prompt your way to that. You earn it. Year after year. Decision after decision.

So leaders, the assignment is clear:
- Before you ask your team to produce faster, teach them to think deeper.
- Before you buy the next platform, sharpen the question it is meant to answer.
- Before you scale, prioritize.

Because the future of leadership will be human, or it will be irrelevant.

#leadership #career #experience