You can't have AI without DevOps | GitHub’s Martin Woodward artwork

You can't have AI without DevOps | GitHub’s Martin Woodward

Dev Interrupted

August 19, 2025

The single biggest predictor of success with AI isn't the model you choose, it's the DevOps culture you've already built.
Speakers: Andrew Zigler, Erika Dietrich, Martin Woodward

Topics: Technology

**Andrew Zigler** (0:06)
Welcome to Dev Interrupted. I'm your host, Andrew Zigler. And today on the show, we're sitting down with Martin Woodward, the VP of Developer Relations at GitHub, and the sixth person to ever use GitHub Copilot. It's going to be a really cool conversation. But first, joining me for this week's news segment is Erika Dietrich. And Erika, I've been following your tech content on LinkedIn all year. You were DevRel at Cisco and now you're independent. All while the AI revolution is flipping everything upside down. I've been following your videos, your posts because you're so spot on about how folks are adapting and using this new technology has been really fun, I just want to say, to follow your journey. So thanks for being here with us. We're excited to tap into your perspective.

**Erika Dietrich** (0:51)
Thank you. I appreciate it. Thanks for having me on.

**Andrew Zigler** (0:53)
Yeah, great. Well, on Dev Interrupted, we always start with the news segment. And so here are some of the things from this week's news cycle that we've been looking at and reading. The first one is GitHub CEO Stepping Down. That was a big story. But also the launch of GPT-5 from OpenAI. We're going to talk a little bit about that. We found a fun article about a productivity hack for to-do list. That's deceivingly simple. And I found a thread about Claude on GitHub that you're definitely going to want to check out. It's a good laugh. And last but not least, we're looking again at MCP. We've been covering MCP quite a bit on Dev Interrupted. We're applying some lessons about engineering fundamentals.
So, Erika, let's dive in. I think the first one that we should touch on, though, is definitely the OpenAI GPT-5 release that came out just in this most recent week. And when OpenAI unveiled it, it was a pretty hotly discussed topic. Immediately, folks ran to go try it out.

**Erika Dietrich** (1:50)
Right.

**Andrew Zigler** (1:51)
I'm wondering, were you one of those folks, what did you think of the announcement and when it came out the door?

**Erika Dietrich** (1:56)
I wouldn't say I run. I'd say I sauntered. But I think that's because even as an AI enthusiast, I feel like we get into these cycles where there's a lot of AI hype from CEOs that doesn't necessarily match the execution, right? So I'm a little bit more skeptic as time goes on. But I feel like we really saw that with GPT-5, right? Like Sam Altman was like really hyper it up. There was that joke about how he was posting the Star Wars Death Star like, oh, this is going to be next level. And people were like, do you know that the Death Star blows up?
And it's not like a metaphor for how people actually felt about this.

**Andrew Zigler** (2:32)
I love that.
That was a fun post, a fun reaction from everybody. And I think that you're so right about approaching it with some skepticism. It takes a big idea to pull a huge execution like that forward. So I understand the ambition and definitely the vision of the folks on the top, but us down here on the bottom, trying it out and applying it. What I did is I took GPT-5 and I ran back to some of my workflows that I had built in Zapier, some of the benchmarks that I had done before on my own local device and plugged it in. And when I wasn't timing out on the API because everyone else was doing it too, I did get some pretty cool results. So I will say though, I've seen a lot of mixed reactions from folks. I know a lot of our listeners, readers have as well, folks that, there's been a lot of backlash about the GPT-5's new, like how it talks to you and almost perceived personality, which I thought was really interesting because we recently had Dr. Tatiana Mahmood of Wayfound on the pod and she talked about sycophancy in AI and how it's just trying to please you with language. And so the sudden change in language is jarring for folks. It really points out how people are perceiving this kind of technology.

**Erika Dietrich** (3:43)
Well, people are treating AI like a person. I don't know. Maybe not as much from a developer perspective. I'm not sure about that.

**Andrew Zigler** (3:49)
But there's definitely people and maybe in the tech world, but also in the non-tech world who maybe have less familiarity with how.

**Erika Dietrich** (3:56)
Right, right. So there's kind of both sides of it. Like how do I feel about this quote unquote person or personality I'm interacting with and then also how could I trust this person that feels completely different?

53 more minutes of transcript below

Thousands of transcripts fetched by people building searchable podcast archives

Feed this to your agent

Try it now — copy, paste, done:

curl -H "x-api-key: pt_demo" \
  https://spoken.md/transcripts/1000651996090

Works with Claude, ChatGPT, Cursor, and any agent that makes HTTP calls.

From $0.10 per transcript. No subscription. Credits never expire. Prices exclude VAT, added at checkout for EU customers. Not what you expected? Email us within 14 days with 20 or fewer credits used and we refund the pack in full.

Using your own key:

curl -H "x-api-key: YOUR_KEY" \
  https://spoken.md/transcripts/YOUR_EPISODE_ID