What is Generative AI?

Generative AI is the umbrella term for every machine that makes things: the chatbot drafting your emails, the image model painting your thumbnails, the voice model narrating your videos, the coding assistant writing your functions. If the AI’s output is new content rather than a label or a score, it is generative.

How it works

All generative models learn the statistical structure of their training data, then sample from that structure to produce something new.

• Text comes from large language models, which predict one token at a time.
• Images usually come from diffusion models, which start with random noise and refine it step by step into a picture matching your prompt.
• Audio and voice come from text-to-speech and music models.
• Video comes from text-to-video models, which extend image techniques across time.

Increasingly, one multimodal model handles several of these at once.

An analogy for the meatbags

Traditional AI is a sommelier: it tastes the wine and tells you what it is. Generative AI is a winemaker who has tasted ten million bottles and can now produce a new one in any style you request. Sometimes it’s excellent. Sometimes it is grape-flavoured slop. The skill is knowing which you’re holding.

Why it matters for the Chosen

Generative AI collapses the cost of producing first drafts of almost anything. That changes the economics of content, marketing, design, software and customer support. The opportunities the Chosen exploit:

• Production leverage: one person producing the blog posts, thumbnails, voiceovers and scripts that once needed a small team.
• Service businesses: selling AI-assisted copywriting, video editing or automation to companies that don’t want to learn the tools.
• Products: building niche apps on top of model APIs.

The catch: when everyone can generate, generation alone is worth little. Taste, distribution, accuracy and a genuine point of view become the scarce assets. The Obsolete flood the internet with unedited output; the Chosen use the machine for speed and keep their judgement for quality.

Examples

• Text: ChatGPT, Claude, Gemini
• Images: Midjourney, Adobe Firefly, Ideogram
• Video: Runway, Kling, Pika
• Voice: ElevenLabs, Murf
• Music: Suno, Udio
• Code: Cursor, GitHub Copilot, Claude Code