A financial advisor I work with landed a $7,500 project this week analyzing five years of a client's financial data. The kind of project that used to take days of manual work. She used Claude to run the analysis. Her expertise, knowing what to look for, what the numbers actually mean, what to tell the client, is what the fee was for. Now she's training her team to run the same process so it doesn't require her direct time every time.
This is what AI adoption for consultants and service businesses looks like when it works. Not replacing the expert. The expert doing more of what she's worth.
There's a reason this is suddenly possible, and it has to do with something that happened in early June 2026 that most people scrolled past.
A senior OpenAI employee told the Financial Times three words: "Chat is dead." OpenAI is doing the biggest overhaul of ChatGPT since it launched. Not a better chatbot. Something different in kind: AI that doesn't wait to be asked. It takes action on its own. Books things. Updates records. Runs a process from start to finish without a human in the middle of every step. The company that created the chatbot era just declared it over.
The market was already voting. ChatGPT held 76% of global AI traffic a year ago. By May 2026, that was 53%. Claude tripled its share. Gemini nearly did the same.
That shift shows up in real life before it shows up in headlines. I've had the same conversation a dozen times in the past few months. Someone mentions they switched from ChatGPT to Claude and then pauses, like they're waiting to be challenged on it. Most of them didn't go looking for an alternative. They started noticing ChatGPT giving them generic answers to questions that needed nuance. Or someone in their network mentioned Claude and they tried it and never went back. A few found out their firm had quietly moved everyone over as an IT decision nobody announced. The migration isn't a product launch. It's a slow erosion of habit, and it's already well underway.
For someone who built a practice around specialized knowledge, a financial advisor, a consultant, a practitioner of any kind, the chatbot era was mostly background noise. A useful tool, maybe. Something clients asked about. Not something that fundamentally changed how the work got done.
What's coming next is different.
A lawyer I know spent about two hours last week on intake and scheduling that his practice coordinator used to handle before she left. He hasn't replaced her because he's not sure what the role looks like anymore. That two hours is exactly where the new generation of AI tools is landing first: the work that surrounds expertise but doesn't require it. The follow-up, the scheduling, the research compilation, the first-pass drafting. The work that matters but doesn't need a decade of experience to move.
I've seen this across enough clients now to know it's a pattern. One client saw her opt-in rate jump from 43% to 66% and checkout conversions increase five times over after we rebuilt how AI worked inside her business. The expertise didn't change. What changed was how much of her time went to the work only she could do.
The less obvious shift is on the client side. The same tools professionals are starting to use, clients are starting to use too. Questions that used to mean a phone call, basic research, standard process questions, first-pass analysis, are increasingly getting answered before the call ever happens. What clients actually show up for now is the judgment that no tool has. The situation where stakes are real, context is complex, and they need someone accountable in the room.
That's not a threat to expertise. It's a clarification of it.
The professionals getting ahead of this aren't the ones chasing every new tool. They're the ones getting honest about where their expertise actually lives versus where they've been the only option. Those aren't always the same thing. AI is doing a good job of making the difference visible.
The technology moved. Whether your practice has is a different question.