Executive Series: Dallas
AI Isn't Testing Technology. It's Testing Judgment.
Notes from a closed-door conversation with the executives closest to the decisions.
There is a lot being said about AI right now. Most of it is loud: predictions about what is coming next, claims about productivity, new platforms and models, and headlines about transformation, disruption, and the jobs AI will replace.
But sit around a table with people who actually have to make the decisions, and the conversation sounds very different.
That was the interesting part of a recent 3Seven Executive Series dinner over whiskey at a distillery in Dallas-Fort Worth. The group included CEOs from professional services firms, an AI software leader, a managed services COO, an investment banker, a go-to-market executive, a former consulting COO, and a CFO. Different businesses, incentives, and perspectives, but as the conversation moved from AI adoption to economics, sales cycles, and the role of human judgment, a surprisingly consistent picture emerged.
The question isn't whether AI is coming. It's what businesses should actually do about it right now.
An AI pilot is not the same as transformation.
This distinction came up repeatedly.
A company can deploy an AI tool without fundamentally changing how the business operates. That's a pilot. Transformation asks whether AI changes the way work gets done, how decisions are made, how customers are served, how teams are structured, or how the economics of the business work.
The technology may be part of that transformation, but it isn't the transformation itself. And organizations can accumulate pilots without ever creating meaningful business change.
As one executive put it during the evening, "AI is really about testing judgment."
That's one of the more useful ways to think about the current moment. The question isn't simply what AI can automate. It's where human judgment still matters. As more decisions become automated, that question becomes more important, not less.
Nobody has completely figured out who wins.
Another source of hesitation around the table was the market itself. Customers haven't universally settled on a model strategy, hyperscaler, platform, or long-term architecture.
Questions around integration, compliance, governance, risk, vendor durability, and platform direction aren't signs that executives don't believe in AI. They're signs that executives understand the cost of making the wrong bet.
That's one reason so much activity remains in proof-of-concept mode. The appetite is there. The conviction isn't always there yet.
For services firms, AI cuts both ways.
For services organizations, the conversation becomes even more complicated.
Customers increasingly expect AI-enabled services to cost less. The expectation is understandable: if technology allows a provider to do more with fewer people, shouldn't the customer see some of that efficiency?
But the economics aren't that simple. AI infrastructure isn't free. Token costs aren't irrelevant. New systems require implementation, oversight, integration, and ongoing investment.
Service providers are being squeezed from both directions: customers want lower prices because of AI while providers are absorbing new costs because of AI.
The bigger strategic question is whether AI ultimately expands the value of the service relationship or simply puts pressure on the price of it. That is a much more consequential question than whether a services firm should simply "use AI."
The value wasn't agreement. It was different vantage points.
The most useful thing about the evening was that nobody had exactly the same perspective.
A professional services CEO sees AI through delivery and margin. A software leader sees it through product and adoption. An investment banker sees it through valuation and outcomes. A CFO sees it through the customer's willingness to pay. A go-to-market leader sees it through increasingly complicated sales conversations.
Put those perspectives around the same table and the same market starts to look different.
One attendee mentioned afterward that he had only known one other person at the table before the evening began. The value wasn't that everyone walked away agreeing. It was that they had tested their thinking against people operating from completely different vantage points.
This is what executive conversations about AI need more of right now: less certainty and more judgment; less prediction and more discussion about what the numbers actually say; less focus on what the technology can do and more focus on what the business should do with it.
The headlines will continue moving faster than the market. The technology will continue changing. The economics will continue evolving. But the decisions still have to be made by people.
And those decisions are getting more interesting.
This was the first stop in the 3Seven Executive Series. We'll continue sharing selected observations from the conversations taking place around the table as the series continues.
The enthusiasm is real. So is the hesitation.
Nobody around the table was questioning AI's long-term impact. The hesitation was about making a significant investment before the market settles.
Technology is moving quickly enough that today's best solution may look very different six months from now. Platforms and pricing models are changing, new vendors appear almost daily, and executives are increasingly asking a reasonable question: What happens if we make the bet today and the economics or technology look completely different next year?
That uncertainty is showing up in the sales cycle. Interest isn't declining; decisions are simply taking longer. Customers want to understand what is possible, but many aren't yet willing to commit to what it will cost.
Which leads to the more difficult conversation.
The AI math still has to work.
For all the excitement around what AI can do, the economic case isn't always as straightforward as the headlines suggest.
One of the more candid discussions centered on whether AI agents are actually less expensive than the human labor they are intended to augment or replace. Token costs matter. Infrastructure and integration matter. So does the ongoing work required to maintain and update a system that often has a remarkably short shelf life.
A deployment that looks compelling today can require meaningful investment to keep relevant as the underlying technology changes. That doesn't mean the economics won't work. It means the burden of proof is changing.
Executives increasingly want to see a measurable connection between an AI investment and a business outcome, not simply a compelling demo or proof of concept.
The question is becoming much more direct: What did it actually change?
The most important AI buyer isn't in IT.
One of the more consequential shifts discussed that evening had little to do with technology. It was about who is making the decision.
AI initiatives have historically been driven through the CIO or CTO organization. Increasingly, business leaders are asking for the technology, defining the problem, and deciding whether the investment makes sense.
That's a meaningful change. A technology leader asks whether a platform can integrate. A business leader asks whether it will improve the margin, shorten the cycle, increase capacity, improve the customer experience, or change the economics of the business.
Those are different conversations.
The organizations best positioned to adopt AI aren't always the ones with the most sophisticated technology strategy. They're the ones that can most clearly articulate the business problem they're trying to solve.
