Your keyboard might actually be the biggest bottleneck in your creative workflow. I recently saw a fascinating update from an innovator based in Singapore who has completely ditched typing in favor of walking, talking, and letting AI handle the rest.
This creator admits that you might spot him wandering the city, smiling suspiciously at his phone while dressed in black. To the outside observer, he looks like he is having a lively conversation with an imaginary friend, but he is actually performing high-level prompt engineering via voice notes. He started this experiment in 2025 and is doubling down on the strategy for the coming year because he realized that typing filters out too much genius.
⚙️ The Mechanics of Voice-First Prompting
The core philosophy here is that the human voice is the most efficient interface for capturing raw ideas. When we type, we have a natural tendency to self-edit; we worry about structure, spelling, and flow before the idea is fully formed. This friction slows down the transfer of information from the brain to the machine.
The expert explains that he uses his voice to capture thoughts in their original, chaotic form. This includes all the nuance, hesitation, and “noise” that usually gets scrubbed out during the writing process. By using a Large Language Model (LLM) like Gemini, he records long, unstructured rambles and asks the AI to act as a synthesizer. The AI takes the raw audio, transcribes it instantly, and, crucially, restructures it into high-quality prompts, social media posts, or technical solutions. This effectively separates the “ideation” phase (human) from the “structuring” phase (machine), allowing for a much faster workflow.
💡 Insight 1: Capturing the Nuance of Thought
One of the most compelling points this professional makes is about the value of “noise” in our communication. When you type a prompt, you are often forced to be linear and logical, but creativity is rarely linear. By using voice, you can speak in circles, backtrack, emphasize certain words with your tone, and explore tangents without losing momentum.
For example, if you are trying to design a training module, you might not know the exact structure yet. In a text-based workflow, you might stare at a blinking cursor waiting for the outline to appear in your mind. In this voice-first workflow, you can simply say, “I want to teach X, and it’s important because of Y, and oh, make sure we don’t forget Z, and the tone should be like this…” The LLM listens to the stream of consciousness and identifies the connections that you might not have consciously realized were there. It turns the “mess” of your thoughts into a structured goldmine.
📌 Insight 2: The LLM as an Editor, Not Just a Writer
This LinkedIn user highlights a critical shift in how we interact with AI tools like Gemini. Many people use AI to generate ideas from scratch, asking it to “write a post about leadership.” However, this often leads to generic, robotic output. This creator’s method flips the script by providing the AI with a massive amount of original context via voice.
When you ramble for three minutes, you are providing the model with roughly 450 words of highly specific context, personal anecdotes, and unique stylistic preferences. The AI’s job then shifts from “creation” to “curation.” It isn’t guessing what you want to say; it is organizing what you have already said. This results in outputs that sound significantly more human because the raw material actually came from a human. This technique is particularly effective for generating solutions or training topics, as mentioned by the author, because it allows you to dump all the technical constraints and requirements verbally without worrying about formatting.
✅ Insight 3: Mobility as a Productivity Multiplier
The environment plays a massive role in cognitive performance, and this innovator leverages that by taking his work to the streets of Singapore. We often think of prompt engineering as a sedentary activity that requires dual monitors and intense focus. However, this post suggests that decoupling yourself from the desk can actually improve the quality of your output.
Walking is known to stimulate divergent thinking, which is essential for brainstorming and problem solving. By combining physical movement with voice capture, the author creates a feedback loop where the environment stimulates new ideas, and the phone captures them instantly. He notes that this method allows him to generate solutions instantly while on the go. It transforms “dead time”, like commuting or walking, into highly productive “deep work” sessions. You essentially carry your entire editorial team and coding assistant in your pocket, ready to transcribe and execute complex tasks the moment an idea strikes.
Potential Challenges and Nuances
While this workflow is powerful, the original poster humorously acknowledges the primary drawback: looking a bit “crazy” in public. Talking to your phone with intensity can draw unwanted attention, and there is a social barrier to overcome when dictating complex instructions in a crowded coffee shop or on a sidewalk. Additionally, while AI transcription is excellent, it can sometimes struggle with technical jargon or heavy background noise, requiring a quick review phase to ensure the “translation” from speech to prompt remains accurate.
The Future is Noisy
This innovative approach challenges us to rethink our reliance on keyboards. If you want to see exactly how this expert structures his workflow or join the conversation about 2026 habits, I highly recommend reading the full post linked below!