Questions instead of answers: morning check-in with AI

Switch to Deutsch

“How are you feeling this morning?” This simple question, asked by an AI language model called Claude, forms the starting point of a daily ritual that has fundamentally changed my working day. While the discussion about language models usually revolves around their ability to generate texts and answer questions, I have discovered another added value: their role as questioners and active listeners. The morning dialogue with Claude – not as a source of information, but as a partner for reflection – creates a moment of clarity in the midst of the everyday chaos of emails, meetings, and projects. This article describes how this unusual use of language models helps me to shift from reactive mode to conscious action and why this format can be particularly enriching for strategic thinking processes.

When ChatGPT appeared at the end of 2022, I watched with interest how most use cases developed: people asked questions and got answers. This approach never convinced me personally. The risk of hallucinations is too significant, and there are more effective tools for pure information search, such as Perplexity.

What fascinated me, however, was a different use of these language models: I don’t ask the questions; the tool asks me.

This perspective seemed much more valuable to me. As someone who has always thought through the writing process – similar to established reflective practices like Julia Cameron’s Morning Pages—I quickly realized the real value is not in the language model’s answers but in my own responses to its questions.

Thoughts outside the head

There is often creative chaos in my head. Thoughts overlap: project ideas, worries, unanswered questions, and half-finished concepts. This confusion regularly leads to a diffuse dissatisfaction and the feeling of losing myself in trivialities.

When experimenting with different language models, I noticed a clear difference in the way Claude conducted the conversation. Instead of immediately presenting me with a list of five pre-formulated questions, Claude asks a question, waits for my answer, and formulates the next one based on it. The result is an actual dialogue.

This quality of active listening and targeted questioning reminds me of therapeutic conversations or coaching sessions—formats that have already helped me gain clarity in the past.

The morning check-in: the emergence of a practice

Over the months, this observation developed into a routine that has since become an integral part of my day: the morning check-in with Claude—15 to 20 minutes that fundamentally change my approach to the day.

The process is basic: I open Claude in the morning and start with

Claude then asks me questions about my current state – physically, emotionally, and mentally. I answer, Claude asks, reflecting back what he has understood.

The result: the diffuse chaos of thoughts in my head becomes structured. Subtle issues that are bothering me become tangible. By formulating answers, I often discover what is actually bothering me.

The reflection trap – an unexpected insight

A particularly significant moment in these conversations came when Claude asked me this question after a few weeks of regular check-ins:

“An additional question for reflection: Could it be that your deep capacity for reflection sometimes leads to ‘analysis paralysis’? In other words, a state in which reflection replaces action instead of preparing it?”

This question uncovered a gap in my thinking: I enjoy in-depth reflection, yet I don’t consistently implement my insights. Reflection itself already creates a sense of progress—without anything actually changing in my life.

This realization led to an important adjustment to my morning check-in: since then, I have always had Claude ask me about specific next steps at the end.

The voice mode: a different quality of thinking

Another discovery has significantly increased the value of these conversations for me: speaking instead of typing my answers.

Typing is ideal for focused, precise formulation. Speaking, on the other hand, allows a freer flow of thought, which is often more valuable for morning reflection. I use MacWhisper for this – a tool that transcribes my spoken words and puts them into an easily readable form.

Claude handles this more fluent style of language surprisingly competently – even when I break off sentences, restart them, or pause to think. It extracts the essence of my statements, often more clearly than I would formulate them myself.

With Claude’s direct voice function, which will hopefully be available soon, this process would become even more natural. Instead of the current detour via dictation software, a real conversation could take place—a continuous dialog without interruptions in the transcription process. This direct voice interaction would make a substantial difference, especially at morning check-in: I could reflect during a walk without having to rely on a screen. Such an arrangement is much more like the natural conversational situation we know from coaching. For me, this would be the next logical step in the evolution of this reflection practice—away from the text-based interface and toward a seamless exchange of thoughts that further lowers the threshold between thinking and articulating.

The protocol: insights over time

An underestimated element of this practice is the protocol. At the end of each check-in, Claude creates a summary of our conversation, which I save in a continuous file. This file becomes part of the “Project Knowledge” in Claude, so the language model has access to the history of my reflections.

This makes two precious things possible:

• Claude can recognize and address patterns in my thinking.
• I, myself, can look back periodically and follow the development of certain topics.

It is particularly interesting to see, after a few weeks, which issues have dissipated and which persist—an indicator of where deeper work may be needed.

The symptoms of failure

The value of this practice becomes particularly clear when I neglect it for a few days. The symptoms are now unmistakable:

• A diffuse dissatisfaction is spreading.
• Thoughts start to circle again in an unstructured way.
• I get lost in little things and details.
• Frustration rises without me being able to clearly name the reason.
• The common thread for the day is missing.

As soon as I notice these patterns, it’s time for a check-in. Even a short session of 10–15 minutes almost always leads to a noticeable breakthrough.

The subtle avoidance: When we shy away from insights

One challenge with the morning check-in often lies deeper than the purely practical question of time management. Occasionally there is an unconscious resistance to the process itself, because anyone who engages in honest reflection must expect insights that have consequences.

I observe this pattern in myself time and time again: if I start to dwell on little things or get lost in trivialities in the middle of a check-in, this is often a sign of a deeper resistance. I avoid it because I suspect that the actual realization would mean that I would have to take certain unpleasant steps or change my behavior.

This subtle avoidance occurs particularly with topics that we deliberately keep vague because the consequences would be uncomfortable. Recognizing this avoidance strategy is a valuable insight in itself. In fact, it has become clear to me: The areas that I avoid the most are often the very areas where the most important insights are waiting.

Mindset instead of productivity trick

It is important for me to emphasize: this practice is not primarily about increasing productivity or achieving remarkable success. It is about something more fundamental — a different way of navigating through the day. With more awareness. With more intention. With more clarity.

The difference does not manifest itself in WHAT I do, but in HOW I do it. The quality of my presence in my life changes. I react less and act more. My decisions are based less on vague feelings and more on conscious reflection.

For me, the effectiveness of this practice is not measured in ticked-off to-do lists but in more subtle signs: more focused work, more present conversations, and fewer moments of helplessness in front of the screen.

Relevance for strategic thinking

From my conversations with managers, I know that the challenges described are often even more pronounced in management positions. A constant change of topics, tightly scheduled meetings, and permanent availability—all of this systematically prevents the moments of reflection that are essential for clear strategic decisions.

The result is a management style that is characterized more by reactive action than proactive design. Strategic decisions are made in day-to-day operations without the necessary mental space.

A structured morning check-in can make a decisive difference here. It creates a conscious moment of pause before the daily stream of demands begins. It enables clearer prioritization and therefore more informed strategic decisions.

Conceptual classification: On the shoulders of best practices

The morning check-in described above may seem new, but it is part of a tradition of tried-and-tested reflection and thinking practices:

Self-distancing

Cognitive research is familiar with the concept of “self-distancing”— the ability to look at one’s own thoughts from an observational perspective. Studies show that people who think about challenges in a kind of dialogue (rather than in a closed inner monologue) develop significantly clearer solutions.

Therapy as the most frequent use of Generative AI

This reflective use of language models is also in line with a broader trend: a recent Harvard Business Review study (2025) shows that the therapeutic use of language models is now the most common use case of all. People are increasingly using AI systems for personal reflection, emotional support, and structuring their thoughts — exactly what the morning check-in offers. This form of use has overtaken all technical use cases within a year, underlining the profound human need for reflective conversations.

Morning Pages

In writing practice, a similar approach exists in Julia Cameron’s “Morning Pages“—uncensored writing in the morning to make unconscious thoughts accessible. The key difference is that while morning pages are completely unstructured, the morning check-in provides more guidance and focus through targeted questions. The point is not to diminish the benefits of morning pages. I personally have had good experiences with them.

Rubber Duck Debugging

The technique of “rubber duck debugging” originates from software development: programmers explain their code to a rubber duck to identify problems. The act of explaining forces precision and often reveals the solution. In our approach, the language model takes on the role of the listener, with the advantage that it can ask intelligent questions.

How questions instead of why questions

The type of question plays a decisive role here: psychology distinguishes between open “how” questions, which lead to solution-oriented thinking, and “why” questions, which often keep us trapped in analytical loops. Language models like Claude seem to intuitively prefer the more productive question types.

Active listening

Last but not least, the approach is based on the basic idea of active listening from Carl Rogers’ talk therapy—a technique that promotes self-awareness through empathetic mirroring and asking open questions without prescribing solutions.

These conceptual parallels explain why the morning check-in works for many people: It combines proven cognitive mechanisms with the specific capabilities of modern language models.

Starting points for your own path

If this approach has piqued your interest, I’d like to share some starting points – not as a strict guide, but as starting points for your own exploration:

1. Just get started: you don’t need a complex setup at first. Open Claude (or another language model of your choice) and say, “I would like to do a morning check-in. Please ask me questions about my current state of mind and my thoughts. Ask one question at a time and wait for my answer.”
2. Find the time: experiment with the duration — sometimes 10 minutes is enough, sometimes it takes 25. The most important question is not how long it takes, but whether you actually get clarity, calmness, and intention into your thoughts.
3. Integrate into routine: the greatest challenge is regularity. My experience: Combine it with an existing morning ritual — the first coffee, breakfast, or the way to the office.
4. For advanced users – create a project: in Claude, you can create a project in which you store detailed instructions. This advantage means you don’t have to enter them every time. You can find details in my project instructions:
   • Communication style (first name, active listening, open questions)
   • Reflection support (identifying emotions, making connections)
   • Logging (summary at the end for later reflection)
5. Explore modes: speaking and typing have different qualities for the thought process. Observe which mode works better for you in which situation.
6. Use self-knowledge: if you have already gained insights into your patterns through coaching or therapy, integrate them.

Differences to other reflection methods

The morning check-in with Claude complements other reflection practices for me:

• Unlike free journaling, asking specific questions helps you to avoid sitting in front of a blank page.
• In contrast to pre-printed reflection diaries with fixed questions, the follow-up questions are personalized and context-related.
• The method can be easily combined with meditation and coaching. It does not replace these but reinforces their effect in everyday life.

Privacy: There are now also options for installing language models locally on your own computer without sending data to the cloud. However, these are not as powerful as Claude but are certainly suitable for basic reflection conversations. Incidentally, Anthropic, the provider of Claude, emphasizes that they do not use conversations to train language models unless the feature is explicitly switched on by the user. With ChatGPT, on the other hand, it is switched on by default but can be switched off in the settings.

Tool for reflection, not a substitute for a relationship

Anyone reading this might be worried about emotional attachment to an AI system or even the replacement of human closeness with digital interaction. I understand that very well.

From my personal experience, however, I can say that, for me, the morning check-in with Claude is in no way comparable to a relationship with a person, be it a therapist, coach, or partner. It is a tool for reflection, comparable to an interactive diary. The use of polite phrases like “good morning” or “thank you” improves the interaction. These communication patterns are anchored in the training data and lead to more helpful responses.

I consciously pay attention to how my relationship with this tool feels, and I keep realizing that I don’t personify it. Claude remains a tool for me—one that helps me create order in my head and structure my thoughts, no more and no less.

Of course, other people’s experiences may differ, and the HBR study indicates that many people actually develop a more emotional relationship with language models. Self-reflection is essential here: how does the interaction feel? Does the tool become a substitute for human relationships, or does it complement them? Being actively aware of these boundaries is an essential part of a healthy approach to technology.

Exploration instead of prescription

Finally, I would like to emphasize that the morning check-in with Claude was a personal exploration for me— a journey of discovery, the experiences of which I share here. There is no “right” or “wrong,” only continuous trial and error and adaptation.

What works for me may be different for others. The essential thing is the experience of using language models not only as an answer giver, but also as a question poser. As support in transforming thoughts from the mind into words – and ultimately into conscious action.

You may discover, as I did, that the most valuable answers don’t come from language models but from yourself—if the right questions are asked.

Have you encountered any issues while experimenting or have any queries regarding the use of Claude? I offer free office hours every Wednesday. Simply book a 15-minute slot.