Just Talk to It
Your voice captures what your keyboard misses — and that difference matters more than you think.
Something Gets Lost When You Type
You know your business better than anyone. You know your customers. You know the nuance of every situation — the context behind the context, the thing that makes this client different from every other client, the reason this situation calls for a specific kind of response.
When you sit down to type a prompt, something happens. You translate. Your brain converts what you know into words, then filters those words for what seems relevant, then types what is left. By the time the prompt reaches the AI, a lot of what you actually know has been left on the floor.
Your voice does not do that. When you talk, you get nuance. You get personality. You get context that you would never think to type because it feels too obvious or too hard to articulate. You get the real story.
That is what stream of consciousness prompting unlocks.
Talk First. Edit Never.
Stream of consciousness prompting is simple: you hit the dictate button on your phone or use voice mode in your AI tool, and you just talk. Not in a structured way. Not following the WWT framework. Just the way you would explain something to a colleague you trust.
You might say something like:
"Okay so I have this client, she has been with me for three years, really loyal, but she just left a two-star review and I am genuinely confused because the appointment went fine, at least I thought it did, she seemed happy when she left, and now I am reading this and she is upset about the wait time but she has never mentioned that before, and I do not want to be defensive but I also feel like the review is a little unfair, and I want to respond in a way that is honest but also does not alienate her because she is a good client and I want her to come back..."
That is not a clean prompt. But it contains everything the AI needs to give you a genuinely useful, nuanced, specific response. Things you would never have typed. The AI reads all of it. And it responds to all of it.
Three Ways to Talk to Your AI
The Dictate Button
Every phone has a dictate/microphone button on the keyboard. In your choice of LLM or Claude on your phone, tap the text field, then tap the microphone icon. Talk. The words appear automatically. Hit send.
- 1. Open your AI tool on your phone
- 2. Tap the text box
- 3. Tap the 🎤 microphone key on your keyboard
- 4. Talk naturally — do not edit yourself
- 5. Hit send when you are done
Voice Mode in Your LLM
Your choice of LLM likely has a dedicated voice mode — usually a headphone or waveform icon in the app. This is a full conversation where the AI talks back to you. Think of it as calling the most knowledgeable advisor you have ever had.
- • Thinking through a complex business decision out loud
- • Getting feedback on an idea as you develop it
- • Brainstorming when you do not even know what you want yet
Voice Memo to Prompt
Record a voice memo on your phone, then use your phone's transcription feature to convert it to text. Paste the transcript with: 'The following is a transcribed voice note. Read it and help me…'
- • Capturing a business idea while you are driving
- • Processing a complex situation after a difficult meeting
- • Drafting content in your natural voice
Use Your Voice When...
- You are processing something complex or emotional
- The context is hard to explain in writing
- You are thinking out loud and have not landed yet
- You are driving, walking, or cannot type
- You want the AI to understand your situation before suggesting anything
- You know exactly what you need
- The task is specific and repeatable
- You want a polished, copy-ready output
- You are at your desk with time to think
- You want the AI to execute a task
These are not competing methods. They are different tools for different moments. The best AI users know which one to reach for.
Your First Stream of Consciousness Prompt
Think of something you are dealing with in your business right now. A situation, a decision, or a communication you need to send.
Head over to the dictate-powered builder, hit the microphone, and just speak your prompt out loud. Talk for 60 seconds. Do not structure it. Do not edit yourself. End with: "Based on everything I just told you, help me [what you need]."
Open the Dictate Prompt BuilderTap the microphone, speak naturally, and watch your prompt take shape.
Neither Method Is Enough On Its Own
The framework gives your AI precision.
Stream of consciousness gives your AI truth.
The best prompt engineers use both.
They talk through the situation first — messy, real, full of context. Then they use the framework to shape the actual request. The framework is the blueprint. Your voice is the story behind it. Together, they give the AI everything it needs to actually help you.
— WWT Prompt Engineering Team
Ready to combine both?
Go to the Prompt Makeover