Last month, OpenAI closed the largest private funding round in history. $110 billion. Amazon alone put in $50 billion. Nvidia added $30 billion. The valuation hit $840 billion.

Then, quietly, in the same week, OpenAI admitted something that should have gotten far more attention than it did.

They said AI has not actually penetrated enterprise business processes.

Not "has not fully penetrated." Has not penetrated. Their words. Despite the funding, the headlines, the breathless LinkedIn posts from every CEO who just discovered chatbots, the technology is not landing inside the organizations that need it most.

So what did they do? They called McKinsey. They called BCG. They called Accenture and Capgemini. They launched a formal alliance, paired their own engineers with these consulting firms, and essentially said: we cannot get this adopted without you. Help us figure out how to make AI work inside real businesses.

I have been thinking about this all week. Because if you are a senior professional with deep expertise in a specific industry, this is one of the most important signals you will see this year.

The company building the most powerful AI tools on the planet just told you, in public, that the tools are not the hard part.

The hard part is knowing where they fit.

Anthropic published something a few weeks ago called the AI Fluency Index. They analyzed nearly 10,000 conversations people had with Claude and tracked 24 specific behaviors that define whether someone is using AI well or poorly.

The central finding was uncomfortable. Most people are getting worse at working with AI as the AI gets better. When the output looks polished, when it comes back formatted and confident and professional, people stop questioning it. They become more directive but less evaluative. They tell the AI what to do, accept what comes back, and move on.

Only 30% of users bother to tell the AI how they want it to work before they start. The rest just type a question and hope for the best.

This is happening at every level. Junior analysts, senior directors, C-suite executives. The tool is getting smarter. The people using it are not keeping up.

I see this constantly in my conversations with senior professionals who are building independent practices. Someone will tell me they are "using AI" and what they mean is they asked ChatGPT to rewrite an email last Tuesday. Or they generated a first draft of a proposal and sent it without reading it critically. Or they spent three hours trying to get a tool to do something that would have taken them 20 minutes to do themselves, because they did not know how to frame the problem.

None of that is AI fluency. That is AI tourism.

Real fluency looks different. It looks like a former head of regulatory affairs who understands exactly which parts of a compliance review AI can handle and which parts require human judgment built over 15 years of working with regulators. It looks like a former managing director who knows that the real reason a financial services firm cannot adopt AI is not a technology problem, it is a governance problem, and he has sat in those governance meetings for a decade. It looks like a former partner at a consulting firm who can look at an AI implementation plan from Accenture and say, in about four seconds, "this will never survive first contact with the operating committee."

That kind of knowledge does not come from a prompt engineering course. It comes from doing the work, in the room, for 20 years.

And right now, almost nobody is packaging that knowledge in a way that organizations can buy.

Think about what OpenAI just did. They have the best technology, functionally unlimited capital, and a team of forward-deployed engineers whose entire job is to sit inside enterprises and build AI solutions. And they still could not crack the adoption problem. They needed domain expertise. They needed people who understand how large, complex organizations actually function.

They went to the big four consulting firms because those firms have relationships and scale. But those firms are generalists. They rotate 26-year-old analysts through six-month engagements. They produce beautiful slide decks. And then they leave, and the organization is back where it started, with a pilot that nobody adopted and a strategy document gathering dust on SharePoint.

You know this. You have watched it happen from the inside, probably more than once.

The gap that exists right now, the one that even OpenAI cannot close with money and engineers, is the gap between what AI can technically do and what a specific organization in a specific industry is actually capable of absorbing. Closing that gap requires someone who understands the politics, the regulatory constraints, the legacy systems, the informal power structures, and the dozen reasons why the last three transformation initiatives failed.

That someone is not a technologist. That someone is an expert who also understands the technology well enough to know where it applies.

If you are building an independent practice, or thinking about building one, this is worth paying attention to. Because the positioning opportunity here is not "I help companies use AI." That is too vague, and there will be a thousand people saying it by next quarter.

The positioning opportunity is: "I know this industry from the inside. I have spent decades navigating the specific constraints that make adoption hard here. And I can show you exactly where AI will work, where it will fail, and what needs to change organizationally before any of it sticks."

That is a high-value, hard-to-replicate offer. It is the kind of thing a Fortune 500 company will pay serious money for, because the alternative is spending $2 million on a consulting engagement that produces a PDF nobody reads.

The people who will thrive in this transition are not the ones who know the most about AI. They are the ones who know the most about their industry and have taken the time to understand how AI intersects with it. The combination is rare. It will stay rare for a while. And it is worth a lot more than either piece on its own.

If I could give you one thing to do this week, it would be this. Pick one process in your domain that you know intimately. Something you have done hundreds of times, something where you can feel in your gut when it is going well or going sideways. Sit down with Claude or a similar tool and try to run that process through it. Not a toy version. The real thing, with the real complexity.

Pay attention to where it works. Pay attention to where it falls apart. Write both down.

What you will have at the end is not just a better understanding of AI. You will have the beginning of a point of view that is uniquely yours, grounded in experience that no one else has, about how this technology meets reality in your specific corner of the world.

That point of view is the foundation of an independent practice. And right now, almost nobody in your industry has built one yet.

Talk soon,

Salama

P.S. If this resonated and you want to talk through how your specific expertise maps to this opportunity, I have a handful of 15-minute strategy sessions open each week. No pitch. Just a focused conversation about your situation and your next step. Apply here.