Back to YouTube
Parker RexMay 25, 2025

How I Use AI for Branding and Marketing (ChatGPT Image Generation)

How I use AI for branding & marketing (ChatGPT image generation). Stand out with AI-driven branding and practical tips for engineers.

Show Notes

Parker walks through how he uses AI to bootstrap branding and marketing, turning a boring product category into a standout brand. He covers a practical workflow: from inspiration and brand kits to AI-generated assets and content strategy.

Branding as differentiation

  • A memorable brand can break through in a crowded market (Liquid Death as a cautionary tale: a bold, wild brand in a boring category wins).
  • As software costs drop, branding, design, and product marketing become critical touchpoints that shape perception across the internet.

From inspiration to a repeatable brand system

  • Start by collecting branding cues you like from other brands (Discord, GitHub, etc.) and build a vision board.
  • Combine elements you admire to craft your own distinct look and feel rather than copying.
  • Build out a cohesive brand kit early: logos, typography, colors, and a clear style direction.

Logos and typography: building a scalable identity

  • Use knockout logos for bold, front-facing identity and supplement with standalone logo variants for different placements.
  • Collect and refine font choices; understand how spacing works (kerning, tracking) to keep logotypes visually balanced.
  • Examples referenced include Mona Sands and HuboSands for headings and numbers, with attention to consistent spacing.

Asset generation workflow with AI

  • Create a board of assets you’ll need: logos (multiple versions), card emblems, CTA emblems, etc.
  • Use AI to generate variations and then refine by hand for pixel precision (e.g., pen tool adjustments for logo perfection).
  • Build multiple “frames” or layouts to see how different assets behave in real-world placements (cards, CTAs, hero sections).

Practical design tools and techniques

  • Use Figma to assemble and arrange assets; leverage keyboard tricks like Shift+A to place assets quickly.
  • Remove backgrounds in Figma (or via plugins) to create clean composites.
  • For font identification, leverage font tools (Adobe Fonts integration and WhatTheFont) to quickly identify typefaces from screenshots or UI samples.
  • When exploring palettes and styles, you can desaturate or recolor assets to test versatility.

The brand JSON: a living design token

  • Generate a brand.json to codify styles: mascot components, primary palette, accents, typography rules, and logo variants.
  • This JSON acts as a single source of truth you can reuse for future generations and tooling.
  • You can iterate on the brand.json as you refine the mascot, colors, and assets.

Mascot design and motion in AI

  • Start with a base mascot (in the video, a character called Val) and evolve it through iterations: helmet style, 3D tweaks, pose variations, and sprite sheets.
  • Use AI to create different versions (e.g., a rocket-riding variant) and then test in motion or short video formats.
  • Tools like Sora can generate variations and motion outputs from base designs, helping you expand usage (stickers, videos, etc.) without starting from scratch.

What to generate next with AI

  • Once you have a brand JSON, prompt AI to produce specific assets (e.g., add a gold chain, swap eyes, try different outfits) to quickly branch new visuals.
  • Create multiple character concepts (e.g., Python mascot, observability-focused agent) to broaden your branding reach.
  • Save all iterations into a centralized Figma page or asset library for easy reuse.

Brand impact on content strategy and performance

  • Branding isn’t cosmetic: it directly influences how viewers perceive thumbnails and videos.
  • Parker shares that a thumbnail style change performed differently from the existing, familiar look; recognizable branding helps audience recognition and click-through.
  • Consistency across thumbnails, logos, and on-screen graphics reinforces trust and recall.

Community and ongoing work

  • The concept of a co-pilot (Val) helps keep content and community questions aligned with the brand.
  • The “Vibe with AI” community is highlighted as a space for collaboration, questions, and asset sharing.

Actionable takeaways

  • Start with a bold stance in a boring category: differentiation beats conformity.
  • Build a brand kit early: knockout logos, standalone variants, fonts, colors, and a brand.json.
  • Assemble a storyboard/board of inspired assets (logos, UI elements, CTAs) before generating assets with AI.
  • Use Figma for asset collection and refinement; leverage background removal and precise vector editing for pixel-perfect results.
  • Identify fonts quickly with font tools (Adobe Fonts integration, WhatTheFont) to speed up typography decisions.
  • Create an AI-driven mascot and expand it with iterations and motion assets; keep a library of variants for reuse.
  • Maintain a living brand.json to ensure consistency across future content and campaigns.
  • Track thumbnail performance to understand how branding affects engagement; align future visuals with what works.