AI can dramatically improve performance while people are using it. A more important question for organizations is whether those gains are also strengthening the capabilities of the workforce itself.
Better Performance Is Not the Same as Better Capability
As organizations invest heavily in generative AI, one assumption tends to pass largely unquestioned. When employees begin producing better work, making better decisions or completing complex tasks more efficiently with AI support, it is tempting to conclude that the workforce itself is becoming more capable. Stronger outputs have traditionally been interpreted as evidence of growing expertise. Generative AI complicates that relationship because exceptional performance may increasingly reflect the quality of the collaboration rather than a lasting change in the individual’s underlying capabilities. An employee can produce an outstanding report because AI contributes structure, knowledge or alternative lines of reasoning that would previously have depended entirely on the person’s own expertise. The work improves immediately, yet it does not necessarily follow that the individual has acquired the same capability independently.
This distinction is becoming strategically important as AI moves from isolated experimentation into everyday knowledge work. Organizations are not merely investing in higher productivity over the next quarter. They are making long-term investments in the quality of their workforce and, ultimately, in their future competitive advantage. Whether AI simply amplifies current performance or also contributes to the development of stronger judgement, deeper expertise and more sophisticated decision-making therefore becomes a question that extends well beyond technology strategy. It touches the foundations of organizational capability itself.
Human Performance Intelligence™ was developed to examine precisely these interactions between people, technology and organizational systems. Rather than assuming that stronger performance necessarily reflects stronger human capability, the framework encourages leaders to distinguish between the performance that emerges within a human-AI collaboration and the capabilities that remain with people once that collaboration has ended. A recent paper published in Scientific Reports, Human-generative AI collaboration enhances task performance but undermines human’s intrinsic motivation, provides an illuminating opportunity to examine this distinction in practice.
What Recent Research Shows
Across four experiments involving more than 3,500 participants, the researchers examined how collaboration with ChatGPT influenced both immediate task performance and subsequent independent work. Participants completed professional writing tasks such as drafting emails, preparing reviews and brainstorming ideas, with some collaborating with AI before returning to comparable tasks without technological assistance.
The immediate findings confirm what many organizations are already observing in practice. Participants working alongside AI consistently produced outputs that were judged to be more analytical, more engaging and of higher overall quality than those produced independently. During the collaboration itself, AI functioned as an effective cognitive partner, extending the quality of work that participants were able to produce.
The more interesting finding emerged only after that collaboration ended. When participants returned to comparable tasks without AI support, the earlier performance advantage disappeared. Their independent work was no better than that of participants who had completed every task without AI from the outset. At the same time, researchers observed an intriguing psychological shift. Participants reported a greater sense of ownership over their work, yet also experienced lower intrinsic motivation and increased boredom. Taken together, the findings suggest that AI substantially enhanced performance while it remained part of the interaction, but those gains did not automatically become enduring individual capability once the interaction ended.
The More Interesting Question
The obvious interpretation of these findings would be to ask what happens when AI is removed from the workplace. From an organizational perspective, however, that is probably the least interesting question because AI is unlikely to disappear from knowledge work. The more consequential question concerns the assumptions organizations make about human development while AI remains present.
For decades, organizations have invested in technologies that enabled employees to perform more efficiently without fundamentally participating in the thinking itself. Generative AI represents a different category of technology. It contributes directly to reasoning, analysis, writing and problem-solving, becoming an active participant in many forms of knowledge work. Under these conditions, it becomes increasingly difficult to infer human capability from observable performance alone. Employees may produce exceptional analyses, persuasive recommendations or highly creative solutions while collaborating with AI, yet those outcomes do not necessarily indicate that the underlying cognitive capabilities responsible for producing them have developed to the same degree.
This distinction deserves considerably more attention because it introduces a tension that many organizations have not yet begun to examine explicitly. Improving today’s organizational performance and developing tomorrow’s workforce are closely related ambitions, but they are not necessarily identical. One concerns the quality of current outputs. The other concerns the enduring capabilities people retain, refine and continue to develop over the course of their careers. AI has the potential to strengthen both. The evidence suggests, however, that one should not automatically be taken as evidence of the other.
What This Means for Leaders
Seen from this perspective, evaluating AI exclusively through productivity gains or improvements in output quality provides only a partial account of organizational progress. These indicators remain essential because they demonstrate whether AI is creating immediate business value. They reveal whether employees work more efficiently, whether decisions improve and whether organizations are becoming more productive. They reveal considerably less about whether the capabilities on which future organizational performance depends are developing at the same pace.
This introduces a second conversation that leadership teams should increasingly bring into their AI strategies. Alongside measuring what employees accomplish with AI, organizations should begin asking how professional judgement continues to develop in an environment where cognitive work is increasingly shared with technology. Which capabilities are being genuinely strengthened through collaboration with AI, and which are merely being supplemented while the technology is present? How should learning, coaching and work design evolve to ensure that critical thinking, judgement and independent problem-solving continue to mature rather than gradually being displaced?
These questions are unlikely to produce universal answers because different forms of work require different forms of collaboration between people and AI. In some contexts, maximizing AI support is entirely appropriate because efficiency and quality are the overriding objectives. In others, preserving opportunities for independent reasoning may prove equally important because long-term capability itself represents a strategic asset. The challenge facing organizations is therefore not deciding between human intelligence and artificial intelligence, but designing work in ways that allow each to strengthen the other over time.
A Broader Lesson
No single study can determine how AI will shape human capability over the coming decade, nor should findings from one experimental setting be interpreted beyond what the evidence can reasonably support. What this research does offer is a valuable conceptual distinction that is likely to become increasingly important as AI becomes embedded in everyday professional work. Better organizational performance should not automatically be interpreted as evidence that people themselves are becoming more capable.
Human Performance Intelligence™ seeks to provide a framework for examining precisely these questions. Rather than asking only how AI changes organizational performance, it asks how people, technology and organizational systems interact over time to shape sustainable human capability. Viewed from that perspective, perhaps the more important question facing leadership teams is no longer whether AI is making employees more productive, but whether organizations are becoming equally intentional about developing the distinctly human capabilities that will remain decisive as AI continues to advance. Those organizations that succeed in strengthening both are likely to discover that competitive advantage depends not only on increasingly capable technology, but on an increasingly capable workforce able to use that technology with sound judgement, adaptability and expertise.