Uploaded May 2026 | Updated September 2026, 1 hour ago

---

In this episode, Zack Ebenfeld and Michał Pierzchała talk about what AI actually looks like inside day-to-day mobile development: self-verifying agents, testing flows, QA support, code review, trust, and the shift from writing code to judging it well.

They also get into the parts that make mobile different from web, especially when it comes to feedback loops, simulators, devices, and verification.

Chapters:
00:00:00 Why You Shouldn’t Trust Agents Blindly
00:14:42 The Big Shift in Agentic Development
00:19:50 Using AI as a Coding Partner
00:25:59 Real “Wow” Moments with GPT-5.4 and Codex
00:29:33 RL Loops vs Self-Verifying Agents
00:35:39 Are We Getting Addicted to AI?
00:40:44 When Agents Get Too Clever
00:45:46 Remote Coding from Your Phone
00:50:34 Building an App from the Gym with Expo Agent
00:54:53 Agent-Powered QA in CI
01:00:03 How AI Is Changing Daily Engineering Work
01:11:29 The Future of Memory, Skills, and Agents.md

Check out episode resources on our website 📚 

Catch more React Universe On Air episodes 🎧 callstack.com/podcast

Sign up for our newsletter ✉️ callstack.com/newsletter

Follow us on X 🐦 https://x.com/callstacki