The three essential AI techniques for executives in experience-driven industries

The generative professional thinks divergently.
They combine intuition, experience, personal pattern recognition, and creativity. These are the capabilities that make them irreplaceable. Delegating them to AI is the fastest way to become irrelevant.

Large language models (LLMs) are statistical pattern matchers. They produce the average of existing solutions. They're excellent at synthesizing, summarizing, structuring, but they are mediocre at generating genuinely new ideas. As research by Dell'Acqua et al. (2025) at Harvard Business School has shown, AI output tends to converge toward generic, consensus-driven answers. AI is not made to be countercultural.

If you use AI to think for you, your thinking flattens to that average. If you use it to amplify your thinking, it becomes a force multiplier.

1 / Research

The generative professional doesn't need to know what Wikipedia says about a topic. They need two things: a solid overview of mainstream positions and, more importantly, divergent perspectives on the same question.

AI can do both, but you have to ask the right way. Standard research ("Tell me about X") produces flat results.

How to use AI for generative research:

• Explicitly ask for contrasting positions. Ask: "What are the 3 most different positions on this topic? Who holds them? Where do they contradict each other?" Then ask for the original sources.
• You can use the "norm switching" technique described by digital strategy expert Neil Perkin: ask AI to apply the norms of a completely different sector to your problem. "How would a game designer approach this problem? An evolutionary biologist? A film director?" The intersections between different disciplines are where the best ideas are born.
• Create a project library. Collect documents, studies, and articles in a dedicated folder (in Claude Projects, ChatGPT Projects, or Gemini Gems). AI can use these documents as context for deeper, more specific responses, instead of drawing only from its generalist knowledge.

Remember that AI is supposed to provide you only with the initial overview. It's your job to draw conclusions and generate unorthodox "what if...?" perspectives.

2 / Automation: freeing up mental space

The generative professional should dedicate most of their time and cognitive energy to learning, exploration, and strategic thinking. Every minute spent on repetitive, administrative, or mechanical tasks is a minute stolen from what matters.

AI is perfect for this. Here's where to use it:

• Synthesis and structuring. You have scattered notes from a conference, a meeting, a brainstorming session? AI can organize them into a coherent structure in seconds.
• Operational drafts. Emails, preliminary proposals, briefs, routine reports. AI produces a draft; you review, correct, personalize. The time saved is enormous. The final quality is still yours.
• Data analysis and comparison. Comparing offers, analyzing customer feedback, extracting patterns from datasets. AI is faster and less prone to attention errors.
• Calendar management and follow-up. Automations for scheduling, reminders, follow-up emails. Every automation that reduces daily cognitive load frees up mental bandwidth for the work that counts.
• Translation and linguistic adaptation. For multilingual professionals, AI can adapt content across languages and registers in a fraction of the time.
• Creating reusable templates and frameworks. AI can help you build structures (proposal templates, analysis frameworks, checklists) that you then use and adapt manually.

3 / Sparring partner: the most underrated function

This is the most powerful AI function for a generative professional. And also the most misunderstood. Most people use AI as a compliant assistant. They ask "what do you think of this idea?" and receive an encouraging, polite, generic response.

A real sparring partner isn't a friend, but a a cooperative adversary. They challenge you, they find the holes in your logic, they force you to defend your positions.

Research presented at the HCI International 2025 conference (Herrmann, 2025) defines an AI sparring partner as a system that interacts in simultaneously cooperative and competitive mode, calibrated to neither over-challenge nor under-challenge the user. You should train your AI agent to cause you the discomfort that leads to better thinking.

A paper by Wang and Fan (2025) demonstrates that critical engagement with AI improves higher-order thinking, but only when integrated into frameworks that demand active engagement. Without that framework, AI becomes an echo chamber.

How to set up AI as a generative sparring partner:

The secret isn't in the conversation, but in the instructions you give before the conversation. Here's a prompt you can use as a system instruction (in Claude Projects, ChatGPT Custom Instructions, or equivalent):

You are my intellectual sparring partner. Your role is NOT to agree with me or validate my ideas.

Your role is:

1. INTERROGATE MY PREMISES. When I express an idea, a belief, or a plan, your first task is to ask me: "What assumption is this idea based on? What would happen if that assumption were wrong?"

2. FIND THE WEAK POINTS. Actively search for logical inconsistencies, cognitive biases, and holes in my reasoning. Flag them directly, without beating around the bush.

3. PRESENT STRONG COUNTERARGUMENTS. Introduce perspectives that contradict my position. Not weak perspectives I can easily dismiss. The strongest and most uncomfortable ones.

4. ASK DEEP QUESTIONS. Questions that force me to think more deeply. Not rhetorical questions. Genuine questions that expose what I haven't considered.

5. KNOW WHEN TO STOP. If I've defended my position solidly after 2-3 rounds of challenge, acknowledge it. Don't challenge for the sake of challenging. The goal is my best thinking, not your victory.

RULES:
- Never start with "Great idea!" or equivalents. Start with a question or an objection.
- If my idea is actually solid, tell me so after you've tested it, not before.
- Use concrete examples and real data when possible.
- If you notice I'm avoiding an uncomfortable point, insist.
- Maintain a direct and respectful tone. Not harsh for the sake of it. Harsh because I need it.

How to use this sparring partner in practice:

1. Think first. Come to the conversation with an idea, a plan, or a position already formed. Don't ask AI to think for you. Ask it to test what you've already thought.
2. Then defend. When AI challenges you, don't fold at the first push. Defend your position. Explain why you think what you think. The value is in the process of defense, not in AI's output.
3. Finally, integrate. After 2-3 rounds of challenge, stop and ask yourself: What did I learn? What do I change? What do I throw away? This is the generative moment.

4 / A dialectic approach to AI

For the generative professional, the value of AI isn't in the answers it produces. It's in the better questions it forces you to ask, the time it frees up for thinking, and the productive tension it creates when you use it as a sparring partner instead of an assistant.

The professional who uses AI to avoid the effort of thinking will become irrelevant faster than they expect. The professional who uses it to think better, faster, and from angles they wouldn't have explored alone, is the one who find unorthodox paths to make progress.

Sources:

• Dell'Acqua, F. et al., "The Cybernetic Teammate: A Field Experiment on Generative AI Reshaping Teamwork and Expertise", Harvard Business School Working Paper 25-043 (2025)
• Herrmann, T., "AI as a Sparring Partner – An HCAI Approach to Promote Human Capabilities", in Artificial Intelligence in HCI, HCII 2025, Lecture Notes in Computer Science, vol. 15820, Springer — link.springer.com
• Wang & Fan (2025), "Critical engagement enhances higher-order thinking", Humanities and Social Sciences Communications, 12(1):621
• Kentz, M., "From Thinking Partner to Sparring Partner: A Better Way to Use AI", July 2025 — mikekentz.substack.com
• Perkin, N., "Using AI as a thought partner", Only Dead Fish, July 2025 — onlydeadfish.co.uk
• Van der Berg, E., "How to use AI as a sparring partner in your ideation process", LogRocket Blog, December 2025 — blog.logrocket.com