Nikonomics - The Economics of Small Business 282 - Best of 2025! How Fear and Regulation Impact AI in Finance with David Kaylor
Feb 26, 2026
David Kaylor, who builds digital banking tools at Alchemy Technology and uses AI to automate banking workflows. He discusses why banking fears and regulation slow AI adoption. He compares Gemini, Grok, Claude and ChatGPT for different jobs. He shows nondevelopers can build automations with Replit and Google Apps Script and shares practical Jira and Sheets automation ideas.
AI Snips
Chapters
Transcript
Episode notes
Regulation Makes Banks Slow Followers On AI
- Banking treats unknown actions as forbidden, which slows AI adoption across the industry.
- David Kaylor contrasts banking's cautious regulatory culture with other businesses that proceed unless explicitly blocked, explaining the sector's slow follow-through.
Built A Clicker Game In 30 Minutes On Replit
- David built a clicker monster game on Replit in ~20–30 minutes using the platform's hosted AI agent and no server setup knowledge.
- Replit handled servers and hosting so David could iterate with natural language prompts instead of manual dev ops.
Pick Models For Their Ecosystem Strengths
- Use each model for what it knows best: Gemini for Google ecosystem tasks, Grok for fast scripting, ChatGPT for general thinking, Claude in Replit for coding workflows.
- David rewrote a Python script into Google Apps Script with Gemini and it ran first try, illustrating fit-for-purpose selection.