You don't need more code. You need more business impact. | Thoughtworks' Chris Westerhold

**Andrew Zigler** (0:06)
Welcome to Dev Interrupted. I'm your host, Andrew Zigler.

**Ben Lloyd Pearson** (0:09)
And I'm your host, Ben Lloyd Pearson.

**Andrew Zigler** (0:11)
In the news this week, we're talking about how Atlassian dropped the state of DevEx results. We're looking at a deep dive into how a Google DeepMind engineer builds context for LLMs. And we're looking a little closer at how Cloudflare is charging crawlers for access to the internet. What do you want to talk about first, Ben?

**Ben Lloyd Pearson** (0:30)
I love these stories this week, but the Atlassian one seems pretty big, so let's go ahead and cover their developer survey results.

**Andrew Zigler** (0:37)
All right. So Atlassian, a leader in AI engineering and adoption of these new tools, dropped the recent State of DevEx report. And we have a look at a quick snapshot of some of the statistics that were included. Obviously, AI adoption being on the rise is something that they are widely covered, but they're talking about how AI tools are producing more value. And according to their report, 68% of developers reported a sizable time savings. Like, we're talking about upwards of 10 hours a week using a tool like AI. And that's a pretty big jump from last year's report where 54% of those developers said that they were experiencing those gains. So according to Atlassian's report, you're seeing actually a lot of velocity in how engineers are adopting these tools and using them. You might remember from recent weeks, this actually runs contrary to reports even from Lead Dev where we hear about how engineers are not seeing these kinds of productive values from tools. So definitely really interesting to see the competing of viewpoints of these surveys. I really recommend folks look closer at them to understand the full story. With all of the reduction in time, Atlassian even goes further to quantify this into how much money you can save as an engineer or an engineering leader. In some cases, they are looking at 10 hours of week of inefficiencies if you have a company with 500 developers. You're looking at like $8 million a year in loss value in terms of waste.
Waste, by the way, is going to be a topic we bring up a little bit later today. But Ben, what do you think about this report?

**Ben Lloyd Pearson** (2:06)
Yeah, that 8 million number is pretty stark. I actually kind of think maybe that is all that adopting AI comes down to. Like how much time did it save or how much work, like work hours did it save you versus the amount that it costs you, you know, because that really does seem to be a good way to evaluate it. But just to be clear, so last year, more than half of respondents to the survey said that they had basically had no measurable improvements from AI. Whereas this year, two-thirds are saying they're saving 10 plus hours using AI. Like that's a massive, massive jump. But you know, honestly, like it actually does kind of resonate with me personally, because I am absolutely saving at least 10 hours a week. It's you know, I've actually been thinking about this a lot lately. Like there's some point where, you know, presumably as this technology matures and gets better, we're all going to start reaching a point where we're capable of doing 2, 3x what we used to do before AI. And I've been thinking about like, what is actually like leading that to happen like just personally for me. And I figured out there's kind of like two different categories of ways that it's helping. There's like the stuff that I've always done that is now substantially easier. So, you know, I work a 40-hour week within that 40 hours. Certain number of hours of that work that I've always done is now significantly less than that. But I think the stuff that's actually more important is all the new stuff that I never would have expected to do before AI that I'm now doing in addition to all of the stuff that I already do. You know, for example, like here at Dev Interrupted, we're now analyzing research white papers in a fraction of the time that it used to take us to, you know, sort of sift through all of these in-depth research articles and try to come away with interesting insights that we can incorporate into our content. And because it's so much easier now, we have the capacity to not only do this in the first place, but then also to execute it at a scale that just never would have been possible for us. And I think that second category is where we're all going to start feeling that like 2x, 3x of our impact and our output.