The Verbalized Sampling OS: 16 System Prompts for Gemini 3, GPT-5.1, Claude 4.5 & Grok 4.1
I thought the Verbalized Sampling technique was powerful in October. Then November happened.
“Update: you will get 50% off on V.1 with V.2 content”
Earlier this year, I released the Verbalized Sampling Promptbook, and the response was insane. +80 sales, a viral Medium story with 5M+ impressions, and hundreds of you asking for more.
But while we were mastering that technique on GPT-4 and Claude Sonnet, the world shifted. In a single "Crazy November," we got Gemini 3, ChatGPT 5.1, Claude 4.5 Opus, and Grok 4.1.
These new models have different brains:
- ChatGPT 5.1 thinks in "Recursive Logic Loops."
- Claude 4.5 manages "Long-Horizon State" better than any human.
- Grok 4.1 has a "Real-Time Truth-Seeking" engine connected to X.
- Gemini 3 sees the world through "Multimodal Branching."
So, I went back to the lab. I took the original Stanford Verbalized Sampling (VS) research (released Oct 10) and rebuilt it from scratch for each of these new architectures.
I didn't just port the prompts. I created Operating Systems.
What You Get (The 4x4 Matrix)
This isn't a PDF you read once. This is a suite of 16 System Prompts (installable "personas") that permanently upgrade how your AI works.
You get 4 Domains x 4 Models:
1. Creative Writing OS
- Claude 4.5 "Holodeck Director": Uses XML state to manage a "Story Bible" and simulate parallel narrative universes.
- Grok 4.1 "Reality Distortion": Scans X (Twitter) in real-time to inject current slang, trends, and "vibes" into your fiction.
- ChatGPT 5.1 "Recursive Muse": Uses logic loops to evolve cliché ideas into high-entropy originality.
- Gemini 3 "Story Weaver": Visualizes branching probability trees for plot development.
2. Marketing OS
- Grok 4.1 "Trend Surfer": Validates marketing hooks against live conversations happening right now.
- Claude 4.5 "Brand Architect": Maintains perfect brand voice consistency across 50+ emails.
3. Research & Business OS
- ChatGPT 5.1 "Logic Analyst": Recursive "Deep Thinking" for market analysis and strategy.
- Gemini 3 "Deep Research": Multimodal synthesis of charts, data, and text.
4. Education OS
- Claude 4.5 "Discovery Engine": A structured tutor that tracks your "learning state" over weeks.
- Grok 4.1 "Truth Architect": Unfiltered, data-driven lessons that debunk myths with real-time charts.
What's Inside the Download?
- 📂 16 .md System Prompt Files: Copy/paste code ready for your API, Workbench, or Chat Settings.
- 📚 16 PDF User Guides: "How to Drive" manuals for each agent (e.g., "How to trigger the 'Wildcard Fork' in Claude").
- “you will get 50% off on V.1”
Who Is This For?
- The "Power User": You already bought my first promptbook and want the "Pro" upgrade.
- The "Model Hopper": You use multiple AIs (Claude for writing, Grok for news) and want a consistent "VS" engine across all of them.
- The Early Adopter: You want to be the first to master the "Crazy November" models properly.
Price: $19.99 (The cost of one month of ChatGPT Plus, but yours forever).
Frequently Asked Questions (Read This First)
“Is this an update to the Verbalized Sampling Promptbook?”
No. This is a standalone, technical product.
• The Promptbook (Vol. 1) teaches you the theory and concepts of Verbalized Sampling.
• This OS (Vol. 2) gives you the executable code (System Prompts) to run that technique perfectly on the newest AI architectures.
“I bought Vol. 1. Is it obsolete now?”
Absolutely not. The theory in Vol. 1 is timeless, and the prompts inside it still work excellently. This new product is a “Power-User Add-on” designed specifically for the models released in November (GPT-5.1, Gemini 3, Claude 4.5, Grok 4.1), which require specific coding languages (XML, JSON) to unlock their full potential.
“If I own Vol. 1, do I need this?”
If you want to manually adapt the prompts yourself, no. But if you want install-and-forget System Prompts that are pre-coded to force these specific new models to use Verbalized Sampling, then this is the upgrade for you.
You Will Master CREATIVITY.