Share
Facebook Facebook icon Twitter Twitter icon LinkedIn LinkedIn icon Email
AI-to-become-a-better-thinker-2

Artificial Intelligence

Will you use AI to become a better thinker, or risk letting AI do your thinking?

Published July 9, 2025 in Artificial Intelligence • 10 min read

Professionals who use AI in a disciplined way to compound their cognitive thinking over time will be the long-term winners. Michael Watkins explains how to make AI a cognitive superpower rather than a crutch

The rapid proliferation of large language models (LLMs) has created a classic prisoner’s dilemma in work environments. In the competitive game of business life, knowledge workers and creatives have been given a choice: use AI extensively to maintain a short-term competitive advantage at the cost of what recent research has shown to be damage to cognitive competence, or invest the extra effort required to use AI strategically while preserving one’s cognitive capability but risk falling behind.

As with all prisoners’ dilemma games, the rational short-term choice seems obvious. When your colleague – or a competing, independent contractor – can generate a comprehensive market analysis in 20 minutes using ChatGPT, while it takes you three hours to think through the problem on your own, the immediate pressure is overwhelming. Multiply this across thousands of daily decisions, from strategic planning to client communications, add a long to-do list, and the competitive imperative becomes clear: adopt AI intensively or fall behind.

Yet this individually rational behavior leads to a harmful outcome. Professionals risk gradually weakening the cognitive skills that support their long-term success. This creates what I see as a race to the cognitive bottom. In this downward spiral, everyone uses AI more heavily than is healthy for each person, leading to widespread cognitive decline, despite each individual’s decision to adopt AI seeming reasonable at the local level.

The research mentioned earlier indicates this isn’t just a theoretical issue. Evidence shows that heavy AI users experience noticeable decreases in neural connectivity, impaired memory retention, and a diminished ability for independent analytical thinking – what the researchers call “cognitive debt.”

Puzzled shocked female office worker overloaded with paper work stunned to have deadline for finishing research holds folders keeps hand on head wears round spectacles poses against pink background
The complexity of AI-assisted work continues to increase, making manual alternatives seem impossibly slow

The competitive trap

I hypothesize that the mechanism driving this race follows a pattern in which:

Potential efficiencies generate pressure to utilize AI. Some professionals begin leveraging AI for routine tasks, such as report generation, initial research, and drafting communications. They appear more productive and responsive, creating competitive pressure for others to follow suit.

Use of AI broadens and deepens. As AI adoption spreads, professionals push these tools into more sophisticated tasks: strategic analysis, problem-solving frameworks, and creative ideation. The complexity of AI-assisted work continues to increase, making manual alternatives seem impossibly slow.

Increasing cognitive substitution occurs. Heavy users begin defaulting to AI for tasks they previously handled independently. Thinking becomes a collaborative process with AI leading, rather than AI supporting human reasoning. Decision-making confidence becomes tied to AI validation.

Memory loss and problems with brain The hand puts the last piece of the puzzle
“Throughout this progression, critical cognitive capabilities appear to deteriorate, some directly demonstrated by research and others inferred from the findings:”

The potential for dependence and deterioration

Professionals reach a point where independent analytical work feels uncomfortable or incomplete. They’ve lost confidence in their ability to tackle complex problems without AI assistance, creating psychological and practical dependence.

Throughout this progression, critical cognitive capabilities appear to deteriorate, some directly demonstrated by research and others inferred from the findings:

Analytical depth. The research demonstrated that LLM users showed “significantly reduced neural connectivity,” indicating decreased active mental processing and analytical thinking. This suggests that excessive cognitive offloading weakens the capacity for strategic thinking and the generation of original insights. Consider two investment analysts: one uses AI to screen opportunities rapidly and then relies on AI-generated analysis frameworks. At the same time, the other uses AI for data gathering but independently develops investment theses. Over time, the first analyst can process more deals but may lose the ability to spot non-obvious patterns that create competitive advantage.

Intellectual ownership. The research directly found that participants who relied heavily on LLMs “reported significantly lower perceived ownership of their essays” and showed “impaired ability to accurately recall or quote specific parts of their essays.” This diminished ownership likely reduces motivation for rigorous evaluation that separates exceptional performers from competent professionals. When your strategic recommendations come primarily from AI prompts, you lose the deep conviction necessary to champion difficult decisions through organizational resistance.

Creative originality. While the research found that LLM-generated essays “exhibited less originality, fewer unique n-grams, and greater homogeneity,” this may reflect the limitations of the AI tool itself rather than proving that users become less creative. However, if professionals consistently rely on AI-generated frameworks, they may become conditioned to homogenized thinking patterns, potentially missing breakthrough opportunities that require genuinely novel approaches beyond existing data patterns.

Perspective diversity. Though not directly tested in the research, consistently relying on similar AI prompts may reinforce existing cognitive biases and narrow perspective-taking, precisely when competitive advantage requires seeing what others miss. AI amplifies the patterns in your questioning without necessarily forcing you to examine your underlying assumptions, potentially creating an illusion of thorough analysis while constraining your thinking.

AI Addiction
This uncertainty raises a related concern: could AI usage become psychologically addictive?

The reversibility question

A critical unanswered question is whether cognitive deterioration from AI dependence is easily reversible. The research provides suggestive evidence about persistence: when participants who had been using AI extensively switched to brain-only writing, they showed notably weaker neural engagement compared to those who had been working independently throughout the day. This conditioning effect suggests that heavy AI use may create lasting changes in how we approach cognitive tasks.

However, the research doesn’t establish how long these effects persist or what’s required to reverse them. We don’t know whether returning to independent thinking quickly restores full cognitive capacity or whether rebuilding analytical skills requires extended effort, during which professionals may remain potentially less competitive than both AI-dependent peers and those who have maintained cognitive independence.

This uncertainty raises a related concern: could AI usage become psychologically addictive? The immediate cognitive relief that AI provides – reducing the mental effort required for complex analysis – may create a dependency cycle where independent thinking feels increasingly uncomfortable or incomplete. If professionals lose confidence in their ability to tackle complex problems without AI assistance, the motivation to rebuild those skills diminishes, potentially creating a self-reinforcing cycle of dependence.

These questions highlight the risk of what might be called a cognitive debt trap: professionals who recognize their declining independence may find themselves unable to afford the time and reduced effectiveness required to rebuild their analytical capabilities, particularly in competitive environments where others continue advancing with AI assistance.

The key insight is that proper AI integration requires more skill and effort than simple delegation but creates sustainable competitive advantages that grow over time.

The long-term winning strategy

The professionals who will thrive understand that this prisoner’s dilemma creates an opportunity. While competitors rush toward cognitive dependence, those who invest in disciplined AI integration can build capabilities that compound over time. As a result, they will develop enhanced cognitive capacity through strategic human-AI collaboration.

This approach requires treating AI as a cognitive amplifier rather than a cognitive substitute, using these tools to enhance rather than replace human thinking capabilities. The key insight is that proper AI integration requires more skill and effort than simple delegation but creates sustainable competitive advantages that grow over time.

Here are six approaches to try:

1 – Maintain cognitive primacy. Generate initial analyses, strategic frameworks, and solutions independently before engaging AI assistance. Utilize AI for iterative refinement while maintaining final decision-making authority and responsibility for critical evaluation.

Example: Market entry strategy

Do: Draft your initial market entry strategy, including target segments, competitive positioning, and resource requirements. Then ask AI to identify potential blind spots, challenge your assumptions about local competition, or suggest alternative go-to-market approaches you hadn’t considered.

Don’t: Ask AI to “write a market entry strategy for Southeast Asia” and present the output as your strategic recommendation, even with minor modifications.

2 – Leverage AI for cognitive expansion. Deploy AI to explore blind spots in your thinking systematically. Use it to argue against your strategic assumptions, generate alternative competitive scenarios, or simulate stakeholder perspectives you might overlook. This deliberate cognitive sparring enhances strategic thinking by forcing you to defend and refine your reasoning.

Example: Investment analysis

Do: After developing your investment thesis, ask AI to “play devil’s advocate against this acquisition strategy.” What are the strongest arguments for why this deal could fail? What would a skeptical board member focus on?

Don’t: Ask AI to “confirm why this acquisition makes sense” and use the validation to reinforce your existing bias without genuine critical examination.

3 – Accelerate learning cycles. Use AI as an intelligent tutor to rapidly acquire domain expertise outside your core competencies. Leverage AI to quickly build foundational knowledge, then use that base to ask more sophisticated questions while ensuring genuine understanding over superficial familiarity.

Example: Technology evaluation

Do: Start with “Explain blockchain fundamentals and their implications for supply chain transparency,” then progress to “How would blockchain implementation affect our pharmaceutical supply chain specifically, given our current ERP systems and regulatory requirements?”

Don’t: Jump immediately to “Create a blockchain implementation plan for my pharmaceutical company” without building the foundational understanding necessary to evaluate the AI’s recommendations.

4 – Enhanced pattern recognition. Train AI to help you identify subtle patterns across large datasets, such as market signals, organizational dynamics, or competitive behaviors. Use these AI-detected patterns as starting points for deeper human investigation, combining AI’s pattern detection with human pattern interpretation.

Example: Customer intelligence

Do: Have AI analyze customer feedback patterns to identify emerging themes, then personally interview customers to understand the context, emotions, and business implications behind the data trends.

Don’t: Accept AI’s pattern analysis as definitive strategic guidance without validating insights through direct stakeholder engagement and business context evaluation.

5 – Strategic automation. Use AI to handle administrative and routine analytical tasks, then redirect the saved cognitive resources toward high-value strategic thinking, stakeholder relationship building, and innovative problem-solving. The key is to ensure that automation truly frees up mental capacity rather than creating dependence.

Example: Financial analysis

Do: Use AI to format financial models, generate routine variance reports, and flag unusual data points, then spend the saved time on strategic scenario planning and business model innovation.

Don’t: Have AI create the strategic scenarios themselves while you focus on operational details, reversing the appropriate cognitive division of labor.

6 – Systematic perspective diversification. Deliberately prompt AI to adopt contrarian viewpoints, different cultural perspectives, or alternative industry frameworks. Use these diverse inputs to stress-test your strategic thinking and identify assumptions you hadn’t recognized.

Example: Strategic planning

Do: “Analyze this strategy from the perspective of a European regulator, a venture-backed startup competitor, and a traditional industry incumbent. What would each group see as our biggest vulnerabilities?

Don’t: Only seek AI perspectives that align with your industry background and existing mental models, as they lack critical external viewpoints.

These practices create a compounding effect that separates disciplined AI users from those trapped in the race to the bottom. While AI-dependent professionals hit a capability ceiling—limited by their prompts and the patterns in existing data—cognitively disciplined professionals continue growing, using AI to reach analytical heights impossible for either humans or AI alone. As AI becomes commoditized, cognitive independence becomes the ultimate differentiator.

The question facing every professional is straightforward: Will you use AI to become a better thinker, or risk letting AI do your thinking?

The strategic choice

Cognitive decline from AI dependence isn’t inevitable, but prevention requires decisive action amid uncertainty. While we don’t know if cognitive effects are permanent or reversible, the strategic choice is clear: assume they may be lasting and act accordingly.

The potential costs are severe: loss of analytical depth, creativity, and intellectual ownership. These risks are significant enough that even a moderate probability of persistence justifies preventive measures. The downside of unnecessary cognitive discipline is minimal compared to permanent cognitive impairment.

Moreover, disciplined AI integration offers immediate benefits, regardless of whether it is reversible. Using AI to enhance rather than replace thinking creates competitive advantages while preserving options. If effects prove reversible, disciplined users still gain superior collaboration skills. If persistent, they’ve avoided the trap entirely.

The question facing every professional is straightforward: Will you use AI to become a better thinker, or risk letting AI do your thinking? Given uncertainty about reversibility and potential psychological dependence, treating cognitive independence as irreplaceable is prudent.

This choice will determine individual careers and whether organizations lead or follow. The most successful will recognize that when everyone has access to the same AI capabilities, enhanced human judgment becomes the ultimate differentiator.

Your cognitive capital remains your most valuable professional asset: your ability to think clearly, creatively, and independently. The challenge is learning to grow this asset while leveraging the most powerful thinking tools ever created.

Authors

Michael Watkins - IMD Professor

Michael D. Watkins

Professor of Leadership and Organizational Change at IMD

Michael D Watkins is Professor of Leadership and Organizational Change at IMD, and author of The First 90 Days, Master Your Next Move, Predictable Surprises, and 12 other books on leadership and negotiation. His book, The Six Disciplines of Strategic Thinking, explores how executives can learn to think strategically and lead their organizations into the future. A Thinkers 50-ranked management influencer and recognized expert in his field, his work features in HBR Guides and HBR’s 10 Must Reads on leadership, teams, strategic initiatives, and new managers. Over the past 20 years, he has used his First 90 Days® methodology to help leaders make successful transitions, both in his teaching at IMD, INSEAD, and Harvard Business School, where he gained his PhD in decision sciences, as well as through his private consultancy practice Genesis Advisers. At IMD, he directs the First 90 Days open program for leaders taking on challenging new roles and co-directs the Transition to Business Leadership (TBL) executive program for future enterprise leaders, as well as the Program for Executive Development.

Related

Learn Brain Circuits

Join us for daily exercises focusing on issues from team building to developing an actionable sustainability plan to personal development. Go on - they only take five minutes.
 
Read more 

Explore Leadership

What makes a great leader? Do you need charisma? How do you inspire your team? Our experts offer actionable insights through first-person narratives, behind-the-scenes interviews and The Help Desk.
 
Read more

Join Membership

Log in here to join in the conversation with the I by IMD community. Your subscription grants you access to the quarterly magazine plus daily articles, videos, podcasts and learning exercises.
 
Sign up
X

Log in or register to enjoy the full experience

Explore first person business intelligence from top minds curated for a global executive audience