Mastering AI: Clare Kitching Outlines Four Modes for Leaders to Stay Involved

In a recent LinkedIn post, Clare Kitching discusses the critical need for leaders to strategically engage with Artificial Intelligence, cautioning against its overuse for tasks that still demand human judgment and oversight. Kitching uses a relatable kitchen analogy to break down different levels of AI integration, emphasizing that choosing the right mode for each task is an emerging core leadership skill.

Kitching begins by highlighting a common pitfall: leaders treating AI like a ‘toaster’ when the task actually requires the more involved ‘pan’ of human attention. She posits:

“Too many leaders are using the toaster for work that still needs them at the pan.”

This initial framing sets the stage for a deeper exploration of how leaders can leverage AI without diminishing their own essential contributions.

The Four Modes of AI Engagement for Leaders

Drawing inspiration from a Harvard Business Review article by David Lancefield, Kitching adapts the concepts into four distinct modes, each illustrated with a kitchen task:

1. Leader-Owned: Making Pancakes

This mode, Kitching explains, is for consequential decisions where the leader maintains direct control and judgment. In this scenario, AI serves as a tool to challenge the leader’s initial thoughts, explore alternatives, and identify blind spots, but the ultimate decision remains with the human leader.

“For consequential decisions, form your own view first. Then use AI to challenge it, explore alternatives and show you what you’ve missed. → You own the call.”

Kitching stresses that in this mode, the leader is actively ‘adjusting the heat and deciding when to flip,’ ensuring the most critical aspects are managed personally.

2. Leader-Shaped, AI-Assisted: Using a Mixer

Here, AI handles the mechanics of a task, but the core meaning, direction, and flavour originate from the leader. Kitching uses the example of preparing a team update or an important speech. The leader defines the message and desired impact, while AI assists in organizing and refining the content.

“The meaning needs to be yours,” Kitching emphasizes, reinforcing that the essence of the communication must remain human-centric.

3. AI-Enabled, Human Review: Baking a Cake

In this mode, AI performs the primary processing, but human oversight is crucial for quality control. Kitching suggests this applies to tasks like summarizing customer feedback or meeting transcripts. While AI can condense information, the leader must still review the output to ensure no critical nuances or signals have been missed.

“AI can summarise customer feedback, meetings and employee concerns. Stay close to the original signals by reading the comments, calling the customer and speaking to your team. → Ensure the summary hasn’t left out something that matters.”

This highlights the importance of maintaining a connection to the source material, even when AI provides a summary.

4. AI-Handled, Bounded: Making Toast

This is the most autonomous mode for AI, suitable for specific, lower-risk tasks with clearly defined parameters. Kitching points to examples like formatting recurring reports or routing routine requests. The key here is establishing clear limits and understanding when human intervention is required.

According to Kitching, this level of delegation is designed to free up leaders’ time for more strategic and impactful work that genuinely requires their unique skills.

The Evolving Skill of AI Integration

Clare Kitching concludes her post by posing a direct question to her audience, prompting reflection on their own practices: “So where are you using the toaster for work that still needs you at the pan?” Her analysis underscores that effective AI adoption isn’t about replacing human input but about intelligently augmenting it. By understanding and applying these four modes, leaders can harness the power of AI while safeguarding their essential judgment, voice, and human connection, ultimately enhancing their leadership effectiveness.