Malt's Editorial Collective gathers insights from our exceptional community of freelancers. The views and opinions expressed in this piece are those of the author, and are published to encourage a diversity of perspectives and insights.
Daniel Racz, transformation manager and productivity expert, outlined the key principles for turning AI into a productivity advantage and presented a practical model for structured AI use at our Spring Breakfast Malt event in Berlin. He currently works as a Change Lead at Roche and as a Leadership and Productivity Trainer (read more @ timecoa.ch). Over the course of his career, he has held more than 18 diverse roles across industries, ranging from global project management to founding his own educational technology startup.
Artificial Intelligence is rapidly becoming part of everyday corporate work. But many organisations still misunderstand where its real value lies. AI is often treated as a shortcut to work faster, but without the right approach, that shortcut can easily lead teams in the wrong direction.
Used intentionally, AI can become a powerful productivity lever: helping professionals structure their thinking, generate first drafts, and automate parts of their workflow. But it is not a substitute for expertise, judgment, or complex consulting work.
For leadership teams and knowledge workers alike, the key challenge is learning where AI creates leverage—and where it does not.
The real competitive advantage will not come from using AI more than others, but from structuring work so that AI creates measurable leverage."
Many professionals expect immediate productivity gains from AI. In practice, the results vary.
The problem: AI is highly effective for ideation, structuring information, and generating first drafts. But when used in isolation—one prompt at a time—the time savings are often modest.
The larger gains emerge when AI becomes part of a structured workflow. Automation, scripts, and connected processes can transform small improvements into significant efficiency gains.
The solution: Treat AI not as a one-off assistant but as a component inside a workflow.
Practical tip: Identify which steps of your work require a first draft: outlines, summaries, visuals, or initial text. Let AI handle the starting point and focus your own effort on refinement and decision-making.
AI’s strengths are often misunderstood—and so are its limitations.
The problem: AI models rely on probabilities and patterns. When tasks involve complex reasoning, judgment, or deep expertise, the generated output can require substantial review and correction. In these cases, the quality control effort may exceed the time saved.
The solution: Use AI primarily for supporting tasks, while keeping responsibility for complex thinking with people.
Practical tip: If a task requires strategic judgment, nuanced analysis, or contextual decision-making, use AI to assist with preparation—such as structuring information or exploring ideas—but not to replace the core thinking.
Many professionals experimenting with AI try to improve productivity by applying it directly to their existing tasks.
The problem: People often ask AI to help complete the same work they already do—writing emails, preparing slides, summarising documents—without changing how the work itself is structured.
The result is faster tasks, but only marginal improvements in overall productivity.
The solution: Real productivity gains often come from changing the process, not simply introducing a tool.
Practical tip: Before starting a piece of work, clarify the intended outcome. Then design the simplest workflow that delivers exactly that result—no more, no less.
AI becomes significantly more effective when combined with structured work methods.
The problem: Without clear boundaries, AI tends to generate more material than needed, leading to unnecessary refinement and rework.
The solution: Apply simple productivity principles such as timeboxing and satisficing.
Timeboxing breaks work into defined activities and time units. Within each step, it becomes easier to decide which tasks should be handled by people and which can be supported by AI.
Satisficing ensures the work meets the defined objective without unnecessary perfectionism—similar to delivering increments in a Scrum environment.
Combine timeboxing and satisficing in three simple steps:
1. Define the outcome (satisficing)
Before starting, clarify what “good enough” looks like.
For example: A presentation draft with clear structure and placeholder visuals.
2. Allocate timeboxes (timeboxing)
Break the work into short, focused segments, such as:
15 min – Draft structure (AI-assisted)
20 min – Generate initial content (AI-assisted)
15 min – Refine and adjust key messages (human-led)
10 min – Final polish only if needed
3. Decide where AI accelerates the work
Within each timebox, ask: Can AI generate a first version of this step?
Use AI for drafts, summaries, and visuals, then refine only what matters.
Once the defined outcome is reached, stop and move on. Avoid additional iterations unless they clearly improve the objective or provide a longer-term value.
A structured approach helps avoid random experimentation and ensures that AI is used intentionally.
1. Clarify the task: What outcome is actually needed?
2. Design the workflow: How should the work be done to reach that outcome?
3. Decide where AI can help: Which steps benefit from AI support or automation?
4. Execute intentionally: Is the plan being followed—or has the work drifted back into improvisation?
This sequence prevents a common mistake: jumping straight to using AI without first understanding the task or designing the workflow.
AI alone does not create productivity. But when combined with structured processes and human expertise, it can significantly accelerate meaningful work.
The organisations that benefit most from AI are not those using it most frequently, but those using it most intentionally.
Technology can accelerate work, but it cannot define what meaningful work actually is. The real question organisations and professionals should ask themselves is therefore not: “How can AI help us do more work?”
But rather: “Is our work structured in a way that allows meaningful work to move faster?”
Only when that question is answered does AI truly become a productivity advantage.
Transformation Manager & Productivity Expert
