Beyond the prompt: Why using AI well is, above all, a communication skill

Is knowing how to use artificial intelligence a 100% technical skill? Not quite. And understanding that can completely change the way you work with it. Access is no longer the differentiator. The same tools are available to anyone who wants to use them. Yet the results can vary dramatically from one person to another. The tool is identical, but what each person gets out of it is not.

At INSPYR Global Solutions (IGS), where technology and talent come together every day, we see this clearly: the people who create the most value with AI are not necessarily those who know the most about models or algorithms, but those who are best at explaining what they need.

The tool responds. Communication determines how well it does.

Think about what happens when you ask a new colleague, “Can you review this code?” They will probably ask: What should it do? What’s not working? What have you already tried?

AI works in a similar way, with one important difference: it often doesn’t ask follow-up questions. It simply responds based on the information it has.

Researchers Herbert Clark and Susan Brennan (1991) explained that communication relies on establishing “common ground,” the shared understanding we build with another person throughout a conversation. Between colleagues, much of that context is assumed. With AI, that context does not exist until you put it into words.

This is exactly what Zamfirescu-Pereira and colleagues (2023) found when studying how people with little or no AI experience wrote instructions for AI systems. Many participants wrote as if the tool already understood what they had in mind, tested it only once, and quickly concluded whether it “worked” or not. The challenge was not the technology itself. It was structuring and expressing the idea clearly.

A Harvard Business School study involving 758 consultants from Boston Consulting Group (Dell’Acqua et al., 2023) adds another important insight. For tasks well suited to AI, those who used it completed more work, faster, and with higher quality. However, when the task required careful interpretation of information, participants who relied on AI were less likely to reach the correct answer.

Same tool, opposite outcomes: The difference lay in how people framed the work and evaluated the responses they received.

It is no coincidence that the World Economic Forum’s Future of Jobs Report 2025 identifies analytical thinking as the most valued skill among employers, while AI and big data rank among the fastest-growing skill areas.

The two are deeply connected: AI amplifies what you already know how to think through and communicate.

Practical ways to communicate better with AI (and your team)

• Write as if you were delegating to a brilliant new colleague: They may be highly capable, but they do not know your project, your previous decisions, or the ideas you’ve already ruled out. What you leave unsaid becomes an assumption.
• Explain the “why,” not just the “what”: “Refactor this module” leaves a lot open to interpretation. “Refactor this module because another team will maintain it and needs to understand it without additional documentation” gives AI a clear objective that guides every decision.
• Clarify the problem before asking for the solution: If you cannot explain the problem in two sentences, the problem may not yet be clear. Sometimes the best first prompt is not asking AI to solve the issue, but asking it to help you define it.
• Ask AI to challenge your thinking: Instead of requesting only an answer, ask it to identify weak assumptions, potential risks, or blind spots in your reasoning. This allows you to use AI as a thought partner, not just an executor.
• Be more skeptical when everything sounds perfect: A response that is clear, confident, and persuasive is not necessarily correct. The more important the decision, the more important it is to verify the outcome.

Beyond the prompt: Communicate better to build better

What makes this skill especially valuable is that it extends far beyond the screen.

The same clarity that creates a strong prompt also creates a better ticket, a smoother handoff between teams, and more effective feedback. People who learn to structure their thinking for AI often find themselves communicating more effectively with other people as well.

At IGS, we believe technical expertise and human skills are not competing forces. They reinforce one another. That is why we value professionals who can master the tools while also thinking clearly, communicating precisely, and collaborating effectively to create better solutions.

How has working with AI changed the way you communicate?

And if you’re looking to grow in an environment where technology and human capabilities advance together, explore career opportunities at IGS and discover what we can build together.

References

• Clark, H. H., & Brennan, S. E. (1991). Grounding in communication. In L. B. Resnick, J. M. Levine, & S. D. Teasley (Eds.), Perspectives on socially shared cognition (pp. 127–149). American Psychological Association. https://doi.org/10.1037/10096-006

• Dell’Acqua, F., McFowland, E., III, Mollick, E. R., Lifshitz-Assaf, H., Kellogg, K., Rajendran, S., Krayer, L., Candelon, F., & Lakhani, K. R. (2023). Navigating the jagged technological frontier: Field experimental evidence of the effects of AI on knowledge worker productivity and quality (Harvard Business School Working Paper No. 24-013). Harvard Business School.
• World Economic Forum. (2025). The future of jobs report 2025. https://www.weforum.org/publications/the-future-of-jobs-report-2025/
• Zamfirescu-Pereira, J. D., Wong, R. Y., Hartmann, B., & Yang, Q. (2023). Why Johnny can’t prompt: How non-AI experts try (and fail) to design LLM prompts. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (Article 437, pp. 1–21). Association for Computing Machinery. https://doi.org/10.1145/3544548.3581388