Why AI Adoption Will Fail Just Like Agile Did—Unless We Change | Darryl Wright

Summary

---

Podcast Summary

Scrum Master Toolbox Podcast: Agile Storytelling from the Trenches

Episode: Why AI Adoption Will Fail Just Like Agile Did—Unless We Change | Darryl Wright
Host: Vasco Duarte
Guest: Darryl Wright
Date: October 29, 2025

---

Episode Overview

In this episode, Vasco Duarte and guest Darryl Wright dive into the parallels between widespread AI adoption and the earlier wave of Agile transformations. They critically analyze why both movements have failed to meet inflated expectations and explore the patterns, pitfalls, and mindsets that lead organizations toward disillusionment. The conversation focuses on how Scrum Masters, Agile coaches, and teams can rethink their approach to adopting new tools—whether Agile or AI—to solve fundamental business problems rather than just chasing trends or silver bullets.

---

Key Discussion Points & Insights

1. Misplaced Expectations: AI as the “New Agile”

• Darryl (03:29): Compares the initial hype around Agile to the current hype around AI, warning that both are often seen as instant solutions when they merely expose existing problems.
  • "People are looking to AI to solve their problems... AI is not a silver bullet, just like Agile wasn't." [02:06]
• Organizations frequently react to disappointment by labeling the failed ‘tool’ as dead, similar to “Agile is dead” rhetoric.

2. The Cycle of Hype, Disillusionment, and Blame

• Vasco (04:58): Recounts a local company in Helsinki that’s oscillated between project management and Agile across three waves, now mixing Agile and AI, yet never truly addressing systemic issues.
  • "If you think Agile is dead, then what are you going to do next?" [05:28]
• Many teams adopted Agile hoping it would fix systemic problems (e.g., speed, value, responsiveness), but failed to examine root causes.

3. The Real Problem is the System, Not the Tool

• Both speakers reference W. Edwards Deming’s principle that changing tools without addressing system-level issues is futile.
• Vasco (07:17):
  • "If you don't look at what are the inherent causes coming from the system itself, no tool will ever solve anything." [07:34]
• The challenge lies in communicating this “systemic thinking” in ways that resonate with teams and leaders.

4. Purpose-Driven Transformation

• Darryl (08:17): Critiques adopting Agile or AI “for its own sake” instead of pinpointing real business or workflow problems.
  • "We're adopting AI for AI's sake instead of 'This is the problem we need to solve. Can AI help us solve that problem?'" [08:36]
• Cites a Buddhist analogy: “Don’t look at my finger when I’m pointing at the moon”—focus on the real problem, not the tool.

5. Team-Based Knowledge Work: New Layers of Complexity

• Vasco (09:24): Emphasizes how software development requires unique, team-based coordination not found in traditional individual knowledge work.
  • "Software is an exponent of that problem... we need a different set of tools." [10:34]
• As AI augments teams, the complexity and risks of coordination grow exponentially.

6. Dangers of Automating Broken Processes

• Darryl (11:24): Warns against automating inefficient processes, advocating for improving or redesigning them first.
  • "If you automate it, isn't that just going to lock in a bad process?" [11:31]

7. The Coordination Challenge with AI Agents

• Vasco (12:06): Predicts that as each team member leverages multiple AI agents, teams will struggle even more to maintain a shared understanding of their work and outcomes.
  • "For teams, AI isn't only a benefit, it's also a huge threat for the ability that we have to cognitively understand what we are doing together." [12:32]
• Darryl (12:52): Agrees and highlights the risk of “divergent chaos” if human and AI collaboration isn’t made explicit.

8. Concrete Experiments: Value Stream Mapping & Explicit Handoffs

• Darryl (14:15): Suggests mapping the value stream to clarify which activities are best suited for humans or for AI, and using clear handoffs and policies (e.g., Kanban principles).
  • "If we were to get people to... ask which stages would humans do and which would AIs do, and have clear visualizations, clear handovers..." [14:24]
• Regular review and improvement beats automation for its own sake.

9. Back to Basics: Problem First, Solution Second

• Darryl (15:31): Channels Einstein:
  • "If I had an hour to solve a problem, I'd spend 55 minutes thinking about the problem and five minutes thinking about solutions." [15:37]
• Teams must devote more time to identifying the right problems before leaping to AI, Agile, or any new solution.

---

Notable Quotes & Memorable Moments

• Darryl Wright:

  • "AI is not a silver bullet, just like Agile wasn't. But the problem is that people don't recognize that." [02:04]
  • "Don't look at my finger when I'm pointing at the moon." [08:51]
  • "If you automate it, isn't that just going to lock in a bad process?" [11:31]
  • "If I had an hour to solve a problem, I'd spend 55 minutes thinking about the problem and five minutes thinking about solutions." [15:37]

• Vasco Duarte:

  • "If you don't look at what are the inherent causes coming from the system itself, no tool will ever solve anything." [07:34]
  • "Software is an exponent of that problem... we need a different set of tools." [10:34]
  • "For teams, AI isn't only a benefit, it's also a huge threat for the ability that we have to cognitively understand what we are doing together." [12:32]

---

Timestamps for Key Segments

• 02:06 – Darryl on AI/Agile not being silver bullets
• 05:28 – Vasco on cycles of Agile adoption/disillusionment
• 07:34 – “It’s the system, not the tool” (Deming principle)
• 08:36 – Adopting for purpose, not fashion
• 10:34 – Team-based knowledge work complexity
• 11:31 – Automating broken processes
• 12:32 – Cognitive risks of adding AI to teams
• 14:24 – Value stream mapping as a clarifying tool
• 15:37 – Problem-first mindset (Einstein quote)

---

Actionable Takeaways

• Always diagnose the problem first: Avoid adopting methodologies or technologies blindly; be clear about the problem and whether the tool actually addresses it.
• Value stream mapping: Use it to reveal where AI might help, where human work is essential, and where processes need to be improved or removed altogether.
• Automation follows improvement: Never automate an inefficient process—fix or eliminate it first.
• Explicit coordination: As AI participation increases, make human/AI responsibilities visible and explicit, using proven methods like Kanban policies.
• Reinforce team understanding: Prioritize clear communication, shared mental models, and explicit handoffs to prevent chaos in human-AI collaboration.

---

Episode Tone

The conversation is open, pragmatic, and laced with lived experience. Both Darryl and Vasco balance caution with optimism, sharing philosophy, system thinking, and concrete techniques. They exhibit curiosity and challenge each other, aiming to equip Agile practitioners with practical tools and mindset shifts, rather than offering hype or platitudes.

---

This summary should provide clear insights, memorable moments, and practical recommendations even for listeners who haven’t caught the episode.

Transcript

[A]

Hey there, agile adventurer, just a quick question.

[B]

What if, for the price of a.

[A]

Fancy coffee or half a pizza, you could unlock over 700 hours of the best agile content on the planet? That's audio, video, E courses, books, presentations, all that you can think of. But you can also join live calls with world class practitioners and hang out in a flame war free and AI slop clean slack with the sharpest minds in the game. Oh, and yes, you get direct access to me, Vasko, your Scrum Master Toolbox podcast. No, this is not a drill. It's this Scrum Master Toolbox membership. And it's your unfair advantage in the Agile world. So if you want to know more, go check out scrummastertoolbox.org membership, that's scrummastertoolbox.org Membership. And check out all the goodies we have for you. Do it now. But if you're not doing it now, let's listen to the podcast.

[B]

Hello everybody, welcome to our Wednesday, the Coaching day here on the podcast. And we have with us this week, Darrell Wright. Hey, Darrel, welcome back.

[C]

Hi Vasque, thanks for having me.

[B]

So on Wednesdays, we pick a topic and then we have a coaching conversation. And just to clarify for everybody out there, we're not having a coaching conversation in the sense that I'm coaching Darrel or Daryl is coaching me. This is about us acting as coaches, expressing curiosity and openness to what might come up during the conversation, figuring out experiments and then trying out that experiment in real life. So building a theory, devising an experiment, running the experiment and learning from that, that's the perspective that we're bringing to this conversation. So, Daryl, introduce the topic to us.

[C]

So anybody listening to this podcast in the time that it was recorded would know that the world has gone AI crazy. And one of the things that I'm seeing that's a massive challenge out there at the moment is that people are looking to AI to solve their problems. And they're doing it in the same way that they previously looked to Agile to solve their problems for them. The problem with that is of course, that Agile doesn't solve problems for you. What it does is it shines a light on where your problems are. You, you still need to do the work to solve them. And so that's why that didn't work then and it doesn't work any better now. Because people are looking to AI to solve their problems. AI will show up where those problems are. They still need to do the work to solve it. In the same way, AI is not A silver bullet, just like Agile wasn't. But the problem is that people don't recognize that. And so how do we help people with something that they don't know they need?

[B]

That is a very good question. Okay, can we think of a specific condition situation you have in mind or do we start talking about the overall expectation that somehow there's this magic tool that if you just adopt it, it will somehow solve all of your prompt? Like what, what, what do you have in mind?

[C]

Well, my particular lens that I thought we could explore first is in the nature of us and our role in this as Scrum masters or coaches or, you know, many people got disillusioned with bad Agile, Agile that wasn't done well or they had a bad report about it and as a result they now have an allergic reaction to the word agile.

[B]

And agile is dead is what they're saying.

[C]

Yes, Agile is dead. And of course Agile isn't dead because Agile is about being collaborative and being closer to your customers and focusing on delivering more value sooner. That stuff is never going to be dead. But because Agile has got a bit of a bad name with those people that didn't understand it, they now don't think they need it. And I think there's a real risk of AI going in the same direction. We're hearing all this stuff about AI by slop and so on.

[B]

And it's everywhere, by the way. It's a real problem.

[C]

Everywhere it is.

[B]

Right.

[C]

And so because I think if people don't understand AI and think that it's going to solve their problems and then it doesn't, they have a bad experience, then in another five years, two years, 20 years, I don't know what it's going to be, we're going to have this massive backlash again against people going, you know, AI is dead because it didn't solve my problems. For me, that is dead. Right. And what does that mean for us?

[B]

If I think there's a very serious point there. So I'm going to try to put the problematic in my own words. Let me see if I got it right. So what you're saying is we had a bunch of expectations on Agile, whatever those were, everybody had a different set of expectations and we adopted Agile based on the expectation that we would get what we wanted as we started adopting Agile and we very likely saw that the old problems continued to exist. We, we got, how do you say, what's the English word? We got disillusioned and frustrated and potentially even a little bit angry. And I do see Anger a lot on LinkedIn, for example. And then we started blaming what we thought once was the cure for the problem. And now we say Agile is the problem. And then I don't know what comes next, because that's the interesting part of the conversation. Because if you think Agile is dead, then what are you going to do next? And I have an example to share with you that I think kind of illustrates this problematic. I know of a local company here in Helsinki who's now in their third Agile adoption wave. They did Agile adoption early in the 2000s, then they moved back to project management. They started in project management, did the first Agile adoption, moved back to project management, then moved back to Agile, then moved back to project management, and now they're in their third Agile adoption, which of course comes mixed with AI, which is an interesting kind of nuance in their last Agile adoption phase. And what I'm thinking of here, and let me know what you think, Dara, but what I'm thinking of here is that when we started looking at Agile was to solve a problem, and that problem was described very often, not in all cases, but very often as we're not getting what we want, we're not getting the right amount of software, we're not getting the right type of software, people are too slow, we can't respond to market demands, let's adopt this Agile thing and kind of fix all of it in one go. But we didn't, and this is where I refer back to Deming, which you mentioned yesterday, but we didn't really think about what's in the system that Agile would help solve or change. What I'm thinking is that with AI, just like with Agile, we're adopting AI thinking that we just continue to work as we did before, and then magically this AI will come in and solve all the problems in how we work now so that we can continue to work like we work now and everything will be solved. Of course, anyone who's read Deming like you did and I did, understands that actually, if you don't look at what are the inherent causes coming from the system itself, no tool will ever solve anything. They will just slightly change the system, but not really transform it. And I'm thinking that maybe this is where we are at, like understanding that. Maybe you and me, I mean, we are at this point of understanding that actually it's not the tool, it's the system. But we don't know how to phrase it in a way that people would get it. Is that where we are?

[C]

I think it Is because for many years I think people were trying to do Agile for Agile sake instead of Agile as a way to get whatever it is they're after, better business outcomes. And I think they're doing the same thing now with AI. Let's adopt AI. Why? Well, because it's the latest thing, because everyone else is doing it. We're adopting AI for AI's sake instead of this is the problem we need to solve. Can AI help us solve that problem? And to me, just to bring in another quotable person, the Buddha is reported to have said, don't look at my finger when I'm pointing at the moon. And it's the same thing. It's asking, what's the problem to solve Here I found that such an incredible tool. It cuts through so much noise. You know, even if you're in a crazy meeting and people talking about all this stuff just coming back to hang on, what is the problem we're trying to solve here? And if you can get clear on the problem, then you can say, okay, will Agile or AI or whatever it is help us solve this problem?

[B]

Okay, so I want to pitch a different perspective. So I'm with you so far, right? Like I agree with you, we need to define the problem clearly. Otherwise we're just blind people kind of trying to touch the world around us and trying to figure out what it is. And by the way, philosophers have been saying this for 2000 years at least, so it's not like we're new to this as a species, right? But what I want to put forward to you is that in software, you and I specifically who work in software, we have a different challenge. So everything you said is right, continues to be right, but we have another layer, and that layer is that software is a significantly different endeavor from the endeavors we've had before. Like for example, if you think about knowledge work, knowledge work mostly happens in the head of one person, like writing movies, writing books, coming up with, you know, artwork, whatever. And lately, so let's say in the last 50 years or so we started to have this new kind of knowledge work. Let's call it the team based knowledge work. So knowledge work that happens not in the mind of one person, but in the mind of a large group of people. And we have serious problems in coordinating disability to come together, generate ideas together, work on ideas together. And software is an exponent of that problem. Because software these days only happens, well, not only, but most of it happens in mass sized teams, like a team of five to nine. But it could also be 10 teams of five to nine and so on. And I think that Agile also brought this perspective that wait a minute, software is different. We need a different set of tools, like the daily stand up, the frequent releases, because software isn't like what we did before. What do you think about that? Are we in the need in our industry to create a different perspective over knowledge work as well?

[C]

I agree. And I think that same advance or that same exponential nature of it that software brought, that you just described, I think AI is going to be that again. I think it's going to accelerate it even more. You know, I have, you know, customers who will say to me, oh, we're going to automate the process for blah blah, blah. And I say, but isn't that currently a really bad process? It's broken, it's really inefficient. Yeah, yeah. So if you automate it, isn't that just going to lock in a bad process? Like wouldn't it be better to make the process good before you automate it? Right, but in their mind it's just, no, no, we'll automate it. That'll fix everything.

[B]

So that's a serious problem. And then I see another serious problem, which is that if you imagine a team of five to nine people, so seven plus minus two, and you imagine this team working on software and they can develop pretty large systems with just, you know, five to nine people. But now imagine every one of those five to nine people working with a bunch of AI agents, creating even more software. And now we have exponentially increased the problem of sharing knowledge, sharing where we are together and building something together. In fact, I would say that for teams, AI isn't only a benefit, it's also a huge threat for the ability that we have to cognitively understand what we are doing together.

[C]

Yeah, I agree. Because especially now that we're starting to see the rise of AI agents, if these AIs are going to be acting autonomously, it's even more imperative that we find a way to work together and be on the same page. Otherwise we'll just create more divergent chaos.

[B]

So I'm thinking that one experiment that we could bring to the people we work with. Okay, the automation one is very clear. Right. Improve before automating. Right. Like that's an obvious thing. Hey, could we make the process 10 times faster? You know, do value stream mapping, figure out. Yes, probably. Because usually processes are very inefficient, even the good ones, not just the bad ones, and automate only after you improve the process or even Remove the need for the process, which is much faster and more productive than automation. But then when we think about teams working with AI, I wonder what we could bring to a team for them to grasp the danger, but also the potential of using AI in the work that we do. Because I think that the coordination problem will be 10x once you introduce AI into the coding and testing and specking. So the specification is also a topic, aspects of the work.

[C]

I think you actually just hinted on something which is value stream mapping. If we were to get people to do experiments with map out your value stream and then ask which stages of that value stream would humans do and which stage would AIs do? And have clear visualizations, clear handovers, the old Kanban explicit policies for moving something from one column to another, and one column is for humans and one column is for AIs. Then we would have a much higher level of visibility of who's doing what and what are the criteria for each one to do. And we would have the opportunity for people to make sure that step by step through the process, someone is making sure that we not having hallucinations, cause advert outcomes because we don't want adverse outcomes. That's something that I'm interested in trying, trying out.

[B]

Absolutely. And this kind of illustrates that. Back to basics, right? It's like define the goal as that was the main point you made. And then like do value stream mapping. Figure out where, you know, maybe there are steps in the process that could be eliminated. That's much more productive than automation.

[C]

Yes. And ultimately I think this one. One of my favorite quotes of all time is from Einstein who said, if I had an hour to solve a problem, I'd spend 55 minutes thinking about the problem and 5 minutes thinking about solutions. And I think people, as human beings, we just often just jump straight to solution mode. I think we need to spend a bit more time thinking about the problems.

[B]

Yeah, I think that's a great way, great way to wrap up this conversation. Thank you for that, Darrell.

[C]

Thanks so much, Fesco.

[A]

All right, I hope you liked this episode, but before you hit next episode, here's the deal. This podcast is powered by people like you. The members who wanted more than just inspiration. They wanted real tools and real connection to people who are practicing agile. Every day we're talking access to over 700 hours of agile gold, CTO level strategy talks, summit keynotes, live workshops, E courses, Deep dive interviews, books, and if you're into no estimates, we got the pioneers of no estimates in those deep Dive interviews. As well Agile business intelligence, creating product visions, coaching your product owner courses, you name it. You'll get invites to monthly live Q&As with agile pioneers and practitioners, plus a private Slack community which is free of.

[B]

All of that AI slop you see everywhere.

[A]

And of course without the flame wars. It's a community of practitioners that want to learn and thrive together. It's the best place to connect with community and learn together. So if this podcast has helped you before, imagine what you will get from this podcast membership. So head on over to scrummastertoolbox.org membership and join the community that's shaping the future of Agile. We have so much for you, so check out all the [e-mail] membership because listening is great.

[B]

It's important.

[A]

But doing it together, that's next level. I'll see you in the community.

[B]

Slack we really hope you liked our show. And if you did, why not rate this podcast on Stitcher or itunes. Share this podcast and let other Scrum Masters know about this valuable resource for their work. Remember that sharing is caring.