How AI is reshaping our work
What’s the real impact of AI on the workforce? Matt Goddard and Nick Warren delve into it, discussing opportunities and challenges as AI reshapes our careers.
As the saying goes, history doesn’t repeat… but it does rhyme!
Watch the video…
Or read the transcript…
Nick: Hey there, friends. I’m Nick Warren, and I’m back with Matt Goddard. You did just say we should label this “Two Old Gits Talk About AI,” but I think we’ve got an interesting perspective. We were young when the last real technological tidal wave came through—specifically, the World Wide Web.
Matt: Yeah.
Nick: I, in particular—and you, to some extent—rode that wave.
Matt: Oh, absolutely.
Nick: We were young, somewhat insulated from the broader effects that hit people who weren’t web-savvy or couldn’t retrain. Whereas now…
Matt: I—
Nick: I won’t say we’re right at the other end of our careers, but…
Matt: What you’re saying is interesting. I’ve just hired a 26-year-old, and I told him AI is going to have a massive impact. It’s either going to improve our efficiencies dramatically, or it’ll start nibbling at our heels on the traditional work we do. We need to get on top of it and understand what’s happening.
Nick: Yeah.
Matt: And because I’m a lot older, I need his unbridled enthusiasm to help me see what I might be missing. A lot of what I see AI do feels like magic.
Nick: Mm-hmm.
Matt: Like we’re in the world of Harry Potter—and I’m here for it. But when I sit down to use it for coding tasks, I think: I know how to code. Why would I tell it to guess what I already know how to do? I can just write it myself and make it exactly right—secure, robust, whatever.
Nick: Hmm.
Matt: I’ve found that AI-generated code might be 60–70% correct. But if I write it, it’s 100% correct because I’m in control of every element. Still, I remember people my age back when the web was emerging saying, “What’s the point of this? I don’t understand it.” And for a long time, the internet was just an add-on to businesses—not the business. Now it’s the default.
Nick: Yeah.
Matt: That’s why I joked about “two old gits” talking about AI—because I can appreciate how amazing these tools are, but I don’t always get it.
Nick: That’s the essence of disruption. It’s Clay Christensen’s idea: the industry being disrupted looks at the new thing and says, “Well, that’s not good enough.” Online shopping used to be poor and insecure. But disruption happens when a new approach is better for the audience than the status quo.
Take me—I haven’t coded in decades. But I was spinning up a Hugo site and used Claude to help me tweak the template. I wouldn’t have attempted that otherwise. That’s the disruptive power right there.
Matt: You’re absolutely right. I’m not some AI techno-bro who thinks AI is the answer to everything. But I’m certain it’s a disruptive force like nothing I’ve seen before. It’s fundamentally changing the way we work and the way businesses operate.
It’s not just about getting rich quick. In the last six months especially, AI tools have shifted from being a quiet productivity boost to being an acknowledged “companion” for delivering work.
It’s like when word processors and spreadsheets became essential office tools. Now, AI is becoming that tool. I don’t even need to open Excel—I can just ask it a question in natural language, like “How many people from the US are in our customer data?” and it answers.
Nick: Yeah.
Matt: It’s more intuitive than I thought it would be. Do you remember early Twitter?
Nick: Mm-hmm.
Matt: It started as status updates—“Matthew Goddard is eating lunch”—but when they removed the prompts, it evolved into a conversation platform. Natural language processing is having the same shift. Before, it was “How old is Judi Dench?” Now it’s “How do I build a willow hurdle?” and it’ll try to work it out with dimensions and all.
Nick: What you’re describing is us getting comfortable using LLMs as search engines—and more. It’s no longer one-off queries, it’s threaded work. We’ve talked before about tools like Make or Gumloop—combining AI with automation to build complex systems.
In content, AI is slightly ahead of where it is with coding. We’re using it to massively increase output. Sure, it can sound average or generic, but the bigger question is: how do you use these tools without losing our voice?
Matt: That’s a really important point. The same is true for coding. When new tech comes along, there’s always a wave of people saying, “Here’s how to get rich using this.” That’s what people are seeing right now—especially developers.
They say, “These so-called vibe programmers don’t understand security.” And that’s valid. You can’t ship vulnerable code. But I’ve also seen AI users evolve. They go from basic “look what I can generate” to learning deeper skills to create secure, robust, and accessible code.
Nick: Mm-hmm.
Matt: That’s the disruption curve. At first, professionals dismiss the early adopters—like how HTML hand-coders sneered at Dreamweaver users back in the day. But over time, those early adopters get better and start doing high-quality work. Eventually, they become the leaders.
Nick: Yeah.
Matt: That shift isn’t coming in five years—it’s happening in months. The change is that fast.
Nick: It’s happening right now. The best quote I’ve seen is: “AI’s not going to take your job. Someone using AI is going to take your job.”
Matt: Yeah.
Nick: I learned that building a business in 1997. You dismiss new tech at your peril. Yes, early AI content can feel average, but the tools are improving fast—look at Claude, look at the latest ChatGPT.
Matt: Totally agree. Back then, we didn’t have the perspective we do now. Now that we’re managing teams, that theoretical 66% reduction in developers has a human face.
Nick: Exactly. One of our responsibilities is to make sure the people we work with understand this shift and are learning the tools. AI makes things easy that weren’t easy before. You can generate 100 blog posts in 10 minutes. But everyone else can too.
The people who succeed will be the ones doing what’s hard. The edge is where the opportunity lies.
Matt: Totally. I was talking to a colleague about how most financial services companies are focused on customer-facing AI, but they’re missing internal tooling. That’s where huge gains are possible—making staff more productive, helping them make better decisions.
And there are two big trends here. First, everyone has access to the same LLMs. So your competitive advantage won’t come from the tool—it’ll come from how you use it. Your domain knowledge, your processes, your curation.
Nick: Yes.
Matt: Second, people with the most experience are best positioned to curate these tools. But that creates a problem: where’s the career path for juniors? If orchestration becomes the key skill, and only seniors can do it, what happens to the next generation?
Nick: Mm-hmm.
Matt: My uncle’s a plumber. Traditionally, you’d pass the business on to an apprentice. But if you build an AI toolset instead, maybe you don’t pass it to a person—you just keep the machine running. If that’s not the case, do we need a guild-like model, where experienced people mentor the next wave?
Nick: I think the people who want to learn will be able to learn better than ever. LLMs make amazing teachers—they focus only on the questions you ask. That’s empowering. But again, it means that motivated self-starters will win.
Matt: Yeah.
Nick: Easy becomes average. To stand out, you still need to do hard things, ask better questions, develop expertise. And for the first time in history, students can go farther, faster. But yes, the old, safe “junior to senior” career ladder may not exist anymore.
Matt: Yeah, and that’s a challenge. It technology creates social mobility—up to a point. You and I have talked about accelerated digital delivery for years. When we started this podcast, we were writing about it back in October 2024. In that article, we laid out the principles: trust, clarity, visibility, and consistency. These are the human elements that allow us to collaborate and communicate well, no matter the delivery method.
Nick: Yeah.
Matt: And of course, you still need skill on top of that. You can have great communication and transparency, but if you’re not good at what you do, you won’t deliver quality results. What’s changed since October is that delivering work doesn’t necessarily mean a team anymore. It can be just you and an LLM working together.
Nick: Yeah.
Matt: But the same principles still apply. You still need trust—that what you get from the LLM is accurate. You need visibility—how did it get to that result? You need consistency—which is tricky because LLMs can give you different answers to the same prompt. And you still need clarity—because a vague prompt gets vague output.
Nick: Yes.
Matt: And even when you’re working solo with an AI, the output still has to be handed over at some point—to a colleague, a client, a user. That chain of delivery still exists. What I find interesting is the space between those tools—the handoffs, the automation, the orchestration. Tools like Make are helpful, but there’s still a lot of friction there.
Nick: Yeah.
Matt: Plus, these tools are all proprietary. OpenAI, Anthropic—they want you to stay inside their ecosystem. So getting LLMs to interact with external APIs or systems is still hard. It’s improving with things like agent design, but it’s early days.
Nick: Yeah.
Matt: Also, let’s not forget—all the code these tools produce is coming from a platform we don’t control. OpenAI, Anthropic… they own the means of your labor. If they retrain their models, your outcomes change.
Nick: Hmm.
Matt: I saw a post—I can’t remember if it was real or an April Fool’s joke—saying you could hire a PhD-level AI software developer for $30,000 a year. Equivalent to five human developers.
Nick: Wow. Okay.
Matt: It’s the same logic that led to outsourcing. But at least back then, you could shop around. Now, if everyone’s using OpenAI’s “PhD dev agent,” and they double the price, you’re stuck.
Nick: Yeah.
Matt: So I’ve been thinking about open-source models—like LLaMA—and whether companies should start building internal capabilities now. Not necessarily retraining their own LLMs, but building tooling around them. That gives you some resilience, some independence.
Nick: Your process, your infrastructure.
Matt: Exactly. Because otherwise, you’re building your business on Amazon’s bookstore again.
Nick: Ha, yeah.
Matt: Do you remember that? Waterstones originally hosted their e-commerce site through Amazon. Someone thought it was a smart move. Cheaper than building their own platform. But then Amazon became the entire market—and Waterstones had no control. It’s the same risk with AI now.
Nick: We need to be anti-fragile. If we want to thrive, we need to build resilience into the system. Anyone listening can probably tell we’re still working this out as we go. But if you haven’t started immersing yourself in this technology yet—start now.
Matt: Yeah.
Nick: Even when it’s not great, even when it’s not as good as you are—you need to see what it can do. I spun up a Docker container the other week to look at something called “Claude Computer Use.”
Matt: Yeah.
Nick: It’s wild. You get a little virtual computer with a browser and Office installed. You give it a task, and it behaves like a human user—browsing, copying, pasting, and editing. It’s nowhere near perfect, but it shows you where the world is heading. I’ll be amazed if this isn’t baked into every OS in a couple of years.
Matt: Oh, absolutely.
Nick: So start now. Everyone’s going to do the easy stuff. If we want to build resilient, future-ready businesses, we’ve got to push further, deeper. Like Wayne Gretzky said—skate to where the puck is going.
Matt: And for people our age, that means letting go of “I can do this better myself.” It doesn’t matter if you can. The point is—this is how work is going to be done now.
Nick: Yeah.
Matt: It’s like using electric power tools. Sure, it’s quaint to do everything with a hand plane and a manual drill, but if you’re a professional, you plug in your tools.
Nick: Yeah.
Matt: We need to get over “it’s not good enough yet.” It’s not perfect now—but this is the worst it’s ever going to be.
Nick: Exactly. Great. So go out and give it a try. Thanks, Matt.
Matt: Thank you.
