AI strategies are increasingly measured by what they can deliver today, whether that is hours saved, costs reduced or greater productivity. There is, however, another question revenue leaders need to ask. What happens to the people who are supposed to develop the judgment required to lead tomorrow?
“AI is not only changing how work gets done. It is changing how people learn to do it,” says Joshua Teixidor, a revenue operations executive and GTM strategist. When technology makes inexperienced output look competent, organizations risk weakening one of the most important mechanisms for developing talent: the feedback loop between doing the work, getting it wrong and having someone more experienced explain why.
AI Amplifies Judgment, or the Lack of It
Teixidor has spent more than two decades leading sales and revenue organizations, with experience spanning forecasting, pipeline management, account strategy, leadership development and go-to-market execution.
The implications of AI become particularly clear in sales. Give the same tool to two salespeople and the difference may have little to do with the technology itself. One uses it to send more generic messages to a larger list. The other considers the person, the business and what might actually matter to that buyer, forms a point of view and then uses AI to research faster and create something more relevant.
“The tool did not create the difference between those two people,” Teixidor says. “The second person made a decision before they touched the software.” AI simply makes each person faster at doing what they were already inclined to do.
For revenue organizations, the challenge is that this difference is becoming harder to see. A rushed email once looked rushed and a weak account plan usually looked weak. Now both employees can produce professional-looking work and fill dashboards with activity. The problem is not that less experienced employees can produce better work sooner. The risk is that better-looking output can hide where judgment is still missing.
Why Policy Cannot Replace Coaching
The instinctive response is often to create an AI policy. Organizations establish rules around which tools can be used, where data can go and what requires approval. Those controls matter, particularly around privacy and governance, but they solve a different problem.
“Policy can define boundaries. It cannot create judgment,” Teixidor says. If guidelines leave room for independent thinking, employees who lack judgment can still produce weak work faster. Tighten the rules enough, and organizations risk removing the very judgment they are trying to develop.
That is where coaching becomes critical. Strong sales leaders do not only inspect whether an account plan was completed or a forecast submitted. They examine the reasoning behind it. Why was this account prioritized? What evidence supports this forecast? What assumption is being made? What did the employee believe before using AI?
The account plan was never valuable simply because the document existed. It was valuable because someone had to think about the account. As Teixidor puts it, “Automate the artifact without protecting the thinking behind it and you can keep the deliverable while quietly removing the reason it existed.” For second-line leaders, this means developing managers who can coach judgment rather than simply pass activity and revenue targets down the organization.
Rebuilding the Apprenticeship
Traditionally, a junior employee produces something weak, a more experienced colleague challenges the assumptions and the employee learns from the correction. Repeated over time, those interactions develop judgment.
AI threatens to remove some of the weak work that made those gaps visible. The answer is not to preserve unnecessary work. It is to recognize that when automation removes a learning loop, leadership needs to replace it deliberately. Research published in the Quarterly Journal of Economics found that generative AI assistance produced particularly strong productivity gains among less experienced customer support workers, alongside evidence that it could help workers learn.
Teixidor points, however, to an important difference between work with fast feedback and work where the answer may not become clear for months. In customer support, an employee can often see quickly whether an issue was resolved. Enterprise sales is different.
An account plan created in March may not be judged until September. An unanswered prospecting message offers little insight into why it failed. A deal that slips can have several plausible causes. When feedback is slow and ambiguous, human coaching becomes more important, not less. Someone still needs to challenge the assumption, ask why and tell an employee when the output looks good but the thinking does not hold up.
The Succession Cost of Better Efficiency
The risk may not be visible in this quarter’s results. It appears later, when an organization needs people capable of making decisions without an obvious answer and discovers there are fewer of them than expected.
For Teixidor, that makes this a succession issue. Succession planning is usually treated as a question of who could take the next role. AI adds another: are organizations still creating the experiences that make someone capable of succeeding once they get there? “The people you will need later are supposed to be developing now,” he says.
That makes leadership development part of an organization’s AI strategy. Revenue scaling depends not only on better tools, cleaner processes or more accurate forecasting. It depends on developing people capable of making decisions when the process does not provide an obvious answer. The responsibility is not to slow AI adoption. It is to decide what can be automated, what still requires independent thinking and where learning loops need to be deliberately protected.
What Leaders Should Inspect Now
Teixidor suggests a simple test. Take five pieces of AI-assisted work from inside the organization and remove the names. Instead of grading the polish, grade the thinking. Does the recommendation follow from the evidence? Does the account strategy make sense? Does the forecast reflect what the customer has actually done? Could the employee explain why they made each decision without pointing back to what the tool generated? Then ask them to explain their reasoning.
That conversation may tell a leader more about the organization’s future capability than an AI usage dashboard. The purpose is not to catch people using AI badly. It is to understand where judgment lives, where it is developing and where it may quietly be disappearing.
AI can make today’s workforce faster, make strong performers better and help less experienced employees perform at a higher level sooner. The harder leadership question is whether it is also helping create the judgment and leadership depth the organization will depend on tomorrow. Because the people an AI strategy may be leaving out are not the employees using AI today. They are the people who have not started yet.
Follow Joshua Teixidor on LinkedIn or visit his website.
Source: Brynjolfsson, Erik, Danielle Li, and Lindsey Raymond. “Generative AI at Work.” The Quarterly Journal of Economics, Vol. 140, Issue 2, May 2025, pp. 889–942.
