From Limits to Possibility: How AI Changes the Starting Point of a Decision

In practice, the most interesting thing AI changes for senior leaders isn’t speed , it’s where a decision begins.

I’ve spent much of my career in industrial automation field, where the job has always been to make well-defined applications run reliably, repeatably, and within tightly controlled parameters. Classic automation is brilliant at executing what we already know; what it never did was widen the question. AI does. When a leader can surface options, stress-test assumptions, and reach perspectives they’d never have arrived at alone, the starting point shifts from “what can we realistically do given our limits?” to “what’s actually possible here?” That tends to produce braver, better-considered decisions.

But industrial automation taught me a discipline that matters more than ever: you never let an automated system run a critical process without verification, safety interlocks, and a named human accountable for the outcome. The same principle now applies to AI in the boardroom. The tools that expand the option space also flood it, and the hard part is no longer generating ideas, it is separating genuinely high-value opportunities from the noise. Automation earns its keep when it frees human attention for that judgement, not when it’s used to skip the thinking.

So the leaders getting the most from AI are not automating their decisions. They are using it to move from constraint-led to possibility-led thinking, while keeping human judgement, context, and accountability firmly on the final call exactly as we always have on the plant floor, now in the space where decisions are made.

Institute's Reflection: Human Judgment Determines Which Ones Matter

One of the most valuable insights in this perspective is that AI changes not only the speed of decision-making but the starting point of thinking. Rather than beginning with existing constraints, leaders can explore a wider range of possibilities, challenge assumptions, and consider options that may not have emerged through conventional reasoning alone. In this sense, AI becomes less of an answer provider and more of a partner for exploration.

Research supports this view. Recent studies show that generative AI can improve the quality of solutions for many knowledge-intensive tasks (Dell’Acqua et al., 2023). However, these benefits are not universal. While AI improved performance on tasks within its capabilities, consultants were less likely to reach the correct solution when relying on AI for tasks beyond its capability frontier. The study also found that the most effective users did not simply delegate work to AI. Instead, they integrated AI into their thinking through different patterns of human-AI collaboration, using it to explore, refine, and challenge ideas while remaining responsible for the final judgment.

The contributor's emphasis on verification and accountability is equally important. Leonardi and Leavell (2026) argue that one of the greatest risks of AI is the illusion of certainty. Because AI often presents responses with confidence and coherence, users may stop exploring alternative explanations too early. Effective decision-making, therefore, requires leaders to deliberately keep uncertainty active, continue questioning assumptions, and treat AI as one source of input rather than the final authority.

This perspective also resonates strongly with our Sense-Making Labs. Participants consistently described AI as most valuable when it helped them move beyond familiar thinking patterns, uncover blind spots, and explore new perspectives before making decisions. At the same time, they emphasized that context, experience, ethics, and accountability remain uniquely human responsibilities. In other words, AI may expand what is possible, but sustainable performance depends on leaders who know how to navigate those possibilities with judgment rather than certainty.

References:

• Dell'Acqua, F., McFowland, E., Mollick, E., 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.

• Leonardi, P. M., & Leavell, V. A. (2026). Knowing enough to be dangerous: The problem of artificial certainty for expert authority when using AI for decision-making and planning. Organization Science.