alivirgo/major-ai-skills/use-voice-typing-for-fast-prompts

Use Voice Typing for Rich Context (The 150-WPM Brain Dump) (AI Skill)

Overview

The #1 cause of poor AI outputs is Context Starvation: users type a rushed, 1-sentence prompt ("Write an email to my team about our roadmap changes") because typing on a keyboard is slow and friction-heavy ($~40\text{ words/minute}$).

Human speech flows at 150 to 180 words per minute.

The Voice Dictation Protocol leverages speech-to-text tools (Superwhisper, Mac Dictation, Whisper AI, or iOS Voice) to deliver a 60-second raw brain dump of rich nuances, edge cases, and emotional context without typing fatigue.

Typing Fatigue vs. Voice Context Density

┌─────────────────────────────────────────────────────────────┐
│                 Prompt Context Density                      │
│                                                             │
│  Typed Prompt (40 WPM / 20 Seconds):                        │
│  "Write an email about moving sprint to next week."         │
│  ↳ 10 words, 0 context, generic output                      │
│                                                             │
│  Voice-Dictated Brain Dump (160 WPM / 20 Seconds):          │
│  "Hey draft an email to the frontend team. We have to move  │
│   the sprint deadline to next Thursday because Dave is on   │
│   sick leave and the Stripe webhook tests are taking longer.│
│   Keep it encouraging, mention pizza on launch day."        │
│  ↳ 50+ rich words, complete nuance, flawless output         │
└─────────────────────────────────────────────────────────────┘

The 2-Step Voice Prompting Workflow

┌───────────────────────────────────────────────────────────────────────────┐
│ STEP 1: Speak Freely (Don't worry about punctuation, 'ums', or grammar)   │
│ STEP 2: Append the "Cleanup & Structure" Instruction                      │
└───────────────────────────────────────────────────────────────────────────┘

Master Voice Cleanup Prompt Modifiers

Append this 1-liner to the start or end of any spoken brain dump:

I just dictated the following stream of thought using voice typing. 
Filter out conversational filler ("ums", repetitions, self-corrections) and:
 
1.Extract the core decisions and tasks.
2.Draft the final [DELIVERABLE: e.g. Executive Slack Update / Client Memo].
 
Dictation:
"[SPOKEN TEXT FROM VOICE TYPING]"

Real-World Case Study

Scenario: Walking Out of a 1-on-1 Client Sync

User's 45-Second Voice Dictation (While Walking to Next Meeting)

"Okay AI, I just met with Sarah from Acme. She loves the design preview we sent yesterday, but she's super nervous about mobile load times because 80% of their shoppers are on phones. She also asked if we can push back the final sign-off from Friday to next Tuesday so their VP of Product can review it. Draft a warm follow-up email confirming the Tuesday extension and explaining that our WebP image compression pipeline will keep mobile load times under 1.2 seconds."

AI-Generated Output (Generated in 3 Seconds)

Subject: Acme Mobile Design Sync & Tuesday Timeline Confirmation

Hi Sarah,

Great catching up today! Thrilled to hear you loved the initial design direction.

To follow up on our discussion:

1. Timeline Extended: We've updated the final sign-off date to next Tuesday, August 29th, so your VP of Product can participate in the review.

2. Mobile Speed Guarantee: Mobile performance is our top priority. With our automated WebP compression and CDN edge caching, mobile load times will remain strictly under 1.2 seconds across all devices.

Looking forward to connecting next Tuesday!

Best,

[Your Name]

Summary Best Practices

• Never edit your speech while talking: Speak in continuous stream-of-consciousness; the AI will parse your true intent flawlessly.
• Use keyboard shortcuts: Map a single key (e.g. Fn Fn or Caps Lock) to trigger local Whisper or OS dictation instantly.