I spent the last 12 hours stress-testing Google's Gemini 3 Pro and their new Antigravity IDE on actual projects.

Everyone's hyping the benchmarks, but does it hold up when you need to ship real work.

In this video, I break down the good (automatic Chrome integration, seamless VS Code migration, powerful agent manager), the bad (reliability issues, context window degradation), and whether it's worth jumping ship from Claude Code.

Plus, my take on why Google's positioning in the AI coding space might matter more than the model itself.

Bottom line: Great tech, growing pains, but not ready to replace your daily driver yet. Here's why—and what to watch for as the space evolves.

In this episode, Armando sits down with automation expert Stephen Pope to cut through the hype and get to what actually matters in the AI automation space.

Stephen shares hard-earned lessons from 17+ years in technology, including building and selling a multi-million dollar agency. The conversation covers the real economics of automation work—how to identify problems worth solving, why the simplest solutions often deliver the most value, and how to avoid building complex systems that clients never use.

Key topics include:

• Finding your first automation clients without the typical sales grind
• Why edge cases reveal whether your automation is actually well-designed
• The relationship between content creation and inbound leads (and why it's harder than it looks)
• How to use analytics without getting lost in spreadsheets
• The importance of momentum and knowing which battles to fight as you grow

Stephen gets refreshingly honest about common pitfalls—from clients who want to use you as a puppet to build their ideas, to the trap of perfectionism when your automation already works. He also shares thoughts on how LLMs are changing the content game and why your story matters more than ever.

This isn't about flashy 100-node workflows. It's about building things that actually solve problems and generate returns—the boring, profitable work that keeps businesses running.

In this technical deep-dive, Armando breaks down Retrieval Augmented Generation (RAG) and why businesses can't rely on basic LLM implementations alone. The episode moves from fundamental concepts to advanced optimization strategies, covering everything from context windows to sophisticated reranking techniques.