I’ve been reading quarterly earnings lately. Not something I normally do. But the AI infrastructure numbers caught my attention, and once you start pulling at the thread, a pattern shows up.
The spending is enormous. Record-breaking, actually. Billions per quarter flowing into GPUs, cloud contracts, AI tooling. Every major tech earnings call has AI as the headline. The money is clearly flowing.
And then I looked at the other side. Surveys about deployment readiness. And the numbers don’t match at all.
Something doesn’t add up
The vast majority of companies now have dedicated AI budgets. That part makes sense. But when those same companies are asked if they feel prepared to actually deploy AI? Less than a third say yes.
I kept rereading that. More than four out of five are spending money on AI. Fewer than one in three feel ready to use it.
That gap feels familiar. I remember reading about cloud adoption going through the same pattern. Someone in leadership reads an analyst report, approves a budget, and months later there’s a line item for AI spending with no clear metric for what it’s producing. The purchasing decision runs ahead of the operational reality.
What I think it looks like on the ground
I don’t have direct enterprise consulting experience here. But I’ve been watching enough case studies to notice a recurring story.
A team deploys an AI assistant. Charges start accumulating faster than expected. The output? Hard to pin down. The tool is running, billing is growing, and the actual value is somewhere between “promising” and “we’re not sure yet.”
I suspect this is happening at scale across thousands of organizations right now. The tools exist. The budgets are approved. But extracting real value seems to be a different skill entirely from buying the tool.
Where I think the gap lives
From everything I’ve been reading and from my own experience integrating AI into my workflow, the gap seems to live in three places:
Integration. Most companies probably don’t have clean data pipelines, clear use cases, or the internal expertise to connect AI capabilities to actual workflows. I know from personal experience that just dropping an AI tool into an existing process doesn’t produce magic. You have to redesign the workflow around it.
Measurement. How do you even measure the ROI of an AI deployment? I struggle with this at the individual level. At the organizational level, without measurement there’s no feedback loop, no iteration, and no way to justify continued spending.
Culture. This one hits closest to home. The companies that feel prepared probably have something the rest don’t: leadership that understands the difference between buying a tool and changing how work gets done. I’ve seen this in my own team. The technology is available to everyone. The willingness to actually change how you work is not evenly distributed.
Why I’m paying attention
I find this gap more interesting than the spending numbers themselves. If most organizations are stuck at phase one (procurement) and phase two is operational integration, then the people who can bridge that distance are going to be valuable.
That’s partly why I’m building this blog and the handbook. I don’t have all the answers. I’m going through the process of figuring out how AI actually changes work, one workflow at a time. And I think that practical, ground-level perspective is exactly what’s missing from the conversation.
The spending numbers are impressive. But the number I’m watching is readiness. When that starts climbing, AI spending turns into AI value. Until then, we’re in the gap. And the gap is where the interesting work is.