The AI-native startup: 5 products, 7-figure revenue, 100% AI-written code. | Dan Shipper (Every)

Key Concepts

• AI-first operations
• Reshoring of American jobs through AI
• Claude Code and Gemini CLI for non-coders
• AI agents and automation
• AGI definition based on agent autonomy and profitability
• Context engineering/Knowledge orchestration
• Vibe checks for AI tools
• Compounding engineering
• Allocation economy
• Generalist vs. Specialist in the AI era
• GPT wrappers

1. Hot Takes on AI

• Reshoring American Jobs: Dan believes AI could bring jobs back to the US by making services like in-house counsel or call centers affordable for smaller companies and individuals. AI allows US-based employees to serve more customers efficiently.
• Claude Code for Non-Coders: Claude Code (command-line interface) and similar tools like Gemini CLI are underrated for non-programmers. They allow processing large amounts of text and automating tasks on local computers. Example: Analyzing meeting notes to identify conflict avoidance.
  • Example: Using Claude Code to analyze meeting notes and identify instances where the user subtly avoided conflict.
  • Example: Using Claude Code to analyze War and Peace to identify Tolstoy's character description techniques.
• AGI Definition: AGI is achieved when it becomes economically profitable to run AI agents indefinitely without human intervention. This is likened to a child's development, where increasing autonomy signifies maturity.
• Critique of Common AI Narratives: Dan rejects headlines about AI replacing jobs or diminishing brain function. He argues that AI usage is a skill, and while some skills may be lost, the gains in other areas are significant.

2. Every: An AI-First Company

• Business Model: Every combines a daily newsletter, AI-powered apps, and a consulting arm to help companies adopt AI best practices.
• Team Size: 15 employees.
• AI Operations Lead: Every has a dedicated "Head of AI Operations" who builds prompts and workflows to automate tasks for the team.
• No Manual Coding: Engineers at Every don't handwrite code; they use AI agents to craft requirements and build products.
• Editorial Arm: Uses AI to improve and accelerate publishing.
• Consulting Arm: Helps companies become AI-first by training employees on AI tools and workflows.

3. AI Tool Stack

• ChatGPT (o3): Used for writing, coding, and self-reflection due to its memory and ability to learn user preferences.
• Claude Code: Essential for building and automating tasks, especially for engineers.
• Gemini CLI: Used for building apps due to its power and cost-effectiveness.
• Codex: OpenAI's coding tool for one-off feature development.
• Granola: Used for meeting transcription and note-taking.
• Monologue: Internal version of Super Whisper/Whisper flow for speech-to-text.
• Notion: Used for meeting recordings.

4. Ways of Operating at Every

• AI Operations Lead: Dedicated role to automate repetitive tasks and build AI workflows.
• Compounding Engineering: Each unit of work should make the next unit easier. Example: Creating prompts to automate PRD (Product Requirements Document) writing.
• Multiple AI Agents: Using different AI agents (Claude, Friday, Charlie) with varying perspectives and personalities.
• GitHub for Prompt Storage: Storing prompts in GitHub for team sharing and collaboration.

5. Building Products at Every

• Identifying Expensive Services: Focus on services traditionally expensive and accessible only to large companies or wealthy individuals (e.g., chief of staff, lawyer).
• Using General-Purpose AI Tools: Initially using tools like ChatGPT and Claude to test the viability of automating these services.
• Unbundling into Apps: If successful, unbundling the AI-powered service into a standalone app.
• Internal Testing: Measuring product success by its adoption and usefulness within Every.
• GPT Wrappers: Creating specialized applications that leverage the capabilities of general AI models.

6. Fundraising Strategy

• Sip Seed Round: Raising a smaller amount of capital to maintain control and optionality.
• Focus on Impact: Prioritizing impact and changing how people see the world over building a massive business.
• Creative Playground: Maintaining a creative and experimental environment within the company.

7. Consulting Arm Insights

• Key Offering: Training companies on how to become AI-first, customized to specific team needs.
• CEO Adoption as Predictor: The most significant factor in successful AI adoption is whether the CEO actively uses AI tools.
• Tactics for Encouraging AI Adoption:
  • Sending an "AI-first" memo.
  • Holding weekly meetings to share AI prompts and use cases.
  • Distributing weekly stats on AI usage across the organization.

8. The Allocation Economy

• Shift from Knowledge Economy: Moving from a focus on specialized knowledge to skills in managing and allocating AI resources.
• Valuable Skills: Evaluating talent, vision, taste, and knowing when to dive into details.
• Generalists: Generalists will become more valuable as AI empowers individuals to dabble in various domains.

9. Conclusion

Dan Shipper and Every are at the forefront of AI adoption in business. Their approach emphasizes AI-first operations, continuous automation, and a focus on empowering employees with AI tools. Key takeaways include the importance of CEO adoption, the value of generalist skills, and the potential for AI to reshape the future of work. The "sip seed" fundraising strategy and the focus on building GPT wrappers highlight a unique and innovative approach to company building in the AI era.