In this episode of the Modern Web Podcast, Rob Ocel and Danny Thompson get real about the current state of AI in software development. They talk about what’s working, what’s frustrating, and why so much of the conversation around AI feels repetitive or disconnected from reality. They explore how AI is actually being used day to day—and where it still falls short.

They also discuss the risks of relying too heavily on AI, especially for junior devs, and why understanding fundamentals still matters.

Keypoints from this episode:

- AI fatigue is setting in among developers, with many eager to move beyond hype and explore tools that solve real problems and deliver everyday value.

- Danny built a conversational AI agent named Nexie, trained specifically on Next.js and Vercel documentation, showing how personalized agents can provide focused, meaningful assistance.

- Over-relying on AI can backfire for junior developers. Using it to generate code without understanding the underlying logic undermines learning and growth.

- AI tools are starting to shift how we think about software quality. In some cases, speed and cost-effectiveness may outweigh traditional goals like maintainability.

Chapters

00:00 – Should Juniors Use AI?

01:04 – Real Talk on AI Conferences

03:00 – Building Nexi: A Personal Dev Agent

05:13 – Beyond the Hype: Finding Real AI Use Cases

09:36 – Why Senior Devs Struggle with AI Tools

15:55 – Mentorship, Theory, and Learning the Right Way

21:10 – When to Use AI (and When Not To)

26:50 – The Growing Gap in Dev Skills

30:31 – AI Is Redefining “Good Software”

32:09 – Final Thoughts + Where to Find Us

Follow This Dot Media on Twitter/X: x.com/thisdotmedia

Follow Danny and Rob on Twitter.

Sponsored by This Dot: thisdotlabs.com