1. 👶 We Are Learning
2. 🛠️ Exploration vs Implementation
3. ◀️ Effective AI Exploration Workflows
4. ➡️ Effective AI Implementation Workflows
5. 🤮 We Hate AI Slop
6. 🛣️ The Future of AI
7. 💎 Getting to World-Class
8. 🏋️‍♂️ Don’t Spoil the Mind

👶 We Are Learning

Nobody knows how this plays out. We are early adopters exploring from first principles, experimenting with what works and what doesn't and maturing as a team from what we learn as we go.

🛠️ Exploration vs Implementation

• ◀️ Exploration: Goals is to test, fail and learn fast.
• ➡️ Implementation: Goal is to ship world-class products.

How we think about AI depends entirely on what mode we are in.

In exploration mode we move fast, experiment freely, fail, understand and treat AI as an experimental tool for speed.

In implementation mode we are building for production, quality matters the most and AI usage needs caution, judgment and accountability.

◀️ Effective AI Exploration Workflows

• Architect: You design, AI implements. Keep ownership of the architecture. AI does the heavy lifting.
• Compression: AI summarizes complexity. Use it to understand docs, codebases and errors faster.
• Ideation: AI generates options. Ask for multiple approaches and choose the best.
• Simulation: AI models systems. Test mechanics, ideas and features before implementation.
• Critique: Ask AI to challenge your work. Use it to find weaknesses in your thinking, design or architecture before implementation.

➡️ Effective AI Implementation Workflows

• Review: AI reads your code and catches what you miss. Use it as a second pair of eyes in production work.
• Refactor: AI cleans and improves existing code. Use it to raise quality without rewriting from scratch.
• Debug: AI diagnoses problems. Describe the issue and let it find the root cause faster than you would alone.
• Atomic Implementation: Break work into the smallest fully completable task and let AI execute it. One atomic task at a time, fully done, before moving to the next.