The Tyranny of the Tangent: The Story of Micro-Chat

As the founder of NooSpan, I spend a lot of time thinking about thinking. My original motivation for building this platform came from my work as a university professor, where my day-to-day involves a great deal of research, learning, and exploration.

While using AI as a thinking partner, I quickly ran into a frustrating, recurring problem with traditional linear chat apps that I’m sure many of you have experienced.

The Pain of a "Quick, Clarifying Question"

Whenever I’m deep into a topic—analyzing a methodology, learning a new theory, or exploring a dataset—the AI will inevitably mention a term or concept I need to clarify. In a traditional tool like ChatGPT, the moment I ask that seemingly innocent question, two bad things happen:

1. I immediately disrupt my main flow. My primary line of thought is now on hold. The conversation is forced onto a detour.

2. The AI often gives me a much longer and more detailed answer than I needed. I might just need a quick definition of a term or a short code example, but the LLM frequently returns a multi-paragraph treatise. The "quick" question is never quick.

This isn’t just an annoyance; it’s a fundamental flaw in the linear chat model.

From Detour to Rabbit Hole

Now, imagine this happening multiple times during a single research session. Pretty soon, the main thread is dotted with these forced, long detours. I begin to lose focus. The clean, logical progression of my work is gone, replaced by a random mix of my main topic and numerous side questions, observations, and clarifications. The conversation loses its coherence.

It gets even worse when a tangent proves to be unexpectedly interesting. Down the rabbit hole I go. Many messages back and forth will fly by without me even realizing where I am or what my original task was. My mental clarity is shattered, and I am desparately trying to pick up my original train of though - often in vain.

The Solution: A Side-Chat That Respects Your Focus

This persistent pain is why I invented Micro-Chat.

The concept is simple but powerful: a quick side-chat in a dedicated side bubble that never disrupts the main conversation flow.

When you encounter a word or phrase you need to clarify, or a new idea to test, you simply highlight the relevant text and click "Ask", which opens a Micro-Chat bubble. You will get your answer in a clean, contained space. As soon as you’re done, you close it. Whenever you want to look at your "side quests" again you just click it open in the main chat. Nice and simple!

The result is transformative. Your main conversation is never interrupted by minor detours. It maintains its topical integrity and logical structure, helping you to maintain mental clarity and focus.

The Power of Constraints: Making it "Micro"

When designing this feature, I was keenly aware of the "rabbit hole" problem. A side-chat that becomes its own endless tangent would defeat the entire purpose.

To prevent this, I intentionally built constraints into the feature:

1. Limited Response Length: I limited the length of each AI reply within a Micro-Chat. This ensures you get the concise definition or quick example you asked for, not an unsolicited essay.

2. A 3-Turn Maximum: Each Micro-Chat is limited to a maximum of three back-and-forth turns. This is enough for a question, a follow-up, and a final clarification, but it prevents the "side quest" from becoming the main quest.

These constraints ensure the feature truly lives up to the "micro" in its name and serves its one true purpose: protecting your main focus.

When to Use Micro-Chat vs. When to Branch

This brings us to a crucial point about the NooSpan workflow. The platform provides two distinct tools for two distinct purposes.

• Use Micro-Chat when your "side quests" are minor and limited in scope. It's perfect for quick clarifications, definitions, fact-checking, or a fleeting brainstorm that doesn't need to become a permanent part of your project's structure.

• Use Branching when you want to dive deeply into a sub-topic. If an idea is important enough to deserve its own dedicated exploration, it deserves a Branch. Branching is designed for deep dives, parallel investigations, and building the permanent, logical structure of your work.

Micro-Chat keeps your main thread clean. Branching gives your deep work the organizational home it needs. Together, they create a workspace that finally allows for both focused inquiry and structured exploration.

Richard
Founder of NooSpan

Richard is the founder of NooSpan, an AI thinking workspace designed for structured exploration. He experienced the limitations of linear chat for deep research and was inspired to build a tool that matches how the human mind actually works. You can connect with him on LinkedIn.