Amsterdam 2023

Generative Value Streams: When Code is not a Constraint

Almost overnight we have entered the frenzy period of the Age of AI. Boards, CEOs and technology leaders are scrambling daily to understand the impact. Some are finding this reminiscent to the early days of open source, and the edge it gave to enterprises whose developers adopted open source frameworks quickly. But the productivity benefits of Generative AI will be orders of magnitude larger than that of any programming innovation that came before it. We should now assume we are entering an era where coding is no longer a constraint of software delivery. Organizations continuing on their existing digital transformations path are akin to those plugging leaks in their central steam steam engine while their competition rebuilds the factory around electrification. In this talk I will overview some of the principles behind the shift, and provide guidance on initiatives you need to have up-and-running today in order to stay ahead of this tectonic shift.


Dr. Mik Kersten

CTO, Planview





Okay. The next speaker is someone who should be familiar to, uh, most of us in this community. Uh, Dr. Kersten, uh, wrote the awesome book project to product four and a half years ago, which is being used by so many organizations, change how they think about their software efforts. And he's now currently CTO of Planview. So over the years, he's taught me so much about architecture, which dictates so much of how organizations are wired, which is a topic of the book that Steve, uh, Dr. Steve Spear and I are working on. So, Mick and I were talking earlier this year, and when he told me about a conversation that he had with a mutual friend of ours, I immediately thought that this should be a talk that he should give to this, uh, group. So I asked him if he could talk about his own reflections as a skeptic in Mo about most things, and ponder how things like ai, uh, will change the software profession as we know it. So here's what's going to happen next. Uh, Mick will give his presentation, and I've asked Dr. Steven McGill, VP of Innovation at Stenotype. Uh, one of my, uh, he's one of my favorite research collaborators to provide a rebuttal to a, uh, to a certain statement that was made. And then he, along with three other people, will give a five minute, uh, mini lecture where, uh, we've asked people to either show us something cool they've done with AI or teach us something about ai. So let's start off with Mick.


Hello, everyone. So for my entire career, I've been looking for where the for 10 x productivity improvement and how software is built. And that journey took me through working on the new programming languages, developer collaboration tools, uh, more recently to value stream management and flow. And to my shock and amazement, I think that much more than 10 x productivity gains is actually gonna come from where we're headed with generative AI and large language models. Uh, it's now we're seeing this age start. Now, I think that, you know, this talk will not be about the social implications of this. It'll be how to understand what's happening to help you and your organizations thrive through, I think, the biggest disruption that any of us are gonna see in our entire careers. And in terms of setting up your organization for success for this, I'm gonna outline a few principles around it and how I think you should start adopting some of these technologies and experiment experimenting with 'em today.


So, just for fun, I did actually use Dai from OpenAI and stability AI to generate all the art in the stock in the style of Dutch Masters since we're here in Amsterdam. But, uh, and by the way, this is a, a portrait done in the style of Rembrandt. Uh, so this is Rod Johnson, a a close friend of mine who's now coding back in the 16 hundreds on his laptop. The Rod, actually, the neat thing with Rod is he hasn't been coding quite that long, but he, he is known as an open source luminary by virtue of having, having created the spring framework, a lot of other open source technologies. And I actually still codes on a weekly basis, sometimes a daily basis even. And he is been my go-to person for understanding some of these major shifts that we have in, in programming. In fact, I remember, uh, back in 2009, rod and I were sitting at a cafe in Amsterdam.


He was preparing his keynote for this spring one conference. Uh, and we were reflect back then, remember oh 8 0 9 was the financial crisis. So he was reflecting on how some banks were, and financial institutions were adopting open source very rapidly at that time, even though of course they had major cus uh, uh, budget cuts. But some of them adopted this new way of consuming and evolving software through open source. And those are the ones that I think have actually thrived and succeeded more over the past decade. Now, what I think will actually happen in terms of the large language model adoption generat of AI, will be much more significant than, than that when Rod and I were digging into this, we had a conversation at the start of the year, he actually said he thinks it's the end of what he's been doing, which is, you know, having coding, but as both a hobby, as a profession.


And it'll happen in, in that most five to 10 years. And so I think one of the main constraints, certainly in every company and every product I've worked on, the main constraint has been delivering the code, building the functionality, delivering that digital experience. And all of a sudden I think that constraint will be gone. So large language models, I believe, will actually remove that constraint on software delivery, which is a very, very fundamental change. I think it's actually going to be an order of magnitude bigger than what we've seen with open source with Agile, with DevOps. I do think, by the way, that it's still very important to listen to all the talks at this conference, because if you've not put in place those DevOps practices, you are not structured properly to leverage the change that's coming, uh, with, with generative ai. Uh, but the thing that I'm most concerned with is that some organizations will be properly structured and set up for this, and we'll take this, this move seriously and move very quickly while others won't, because we're gonna get into this very big frenzy.


So again, the goal here is to make sure that you bring this back to your organization, your leadership, and make sure that you lean into this change because, uh, we don't want your organization to get left behind. And I actually see this as one of these major technological revolutions. So the, for those of you who've read Project to Product, it talks about those, these five big technological revolutions, some of the industrial revolution, the age of steam and railways. And, uh, this is really documented by the work of Carl Perez who wrote the book Technological Revolutions and Financial Capital and Interest. Interestingly, in early 2019, I had a call with, with Carlotta, and I asked her, I was reading a bit more about ai, had no idea that we would be where we are today with large language models, but I asked her, you know, what do you think the sixth technological revolution will be?


And I said, you know, it seems like it's going to be AI that probably won't happen for another decade or two, which of course was completely wrong in terms of the, the power that we're seeing now. Uh, and she said, well, one thing we know about these revolutions is before the installation period, uh, you don't know what it's going to be yet. There are too many things being tried and experimented, experimented with, but as soon as something fundamentally changes around the constraint on production, and that's what's always happened, right? Electricity changed the constraint on a factory where Henry Ford was able to all of a sudden not be constrained by a steam engine in the middle of a factory floor, but get power to the whole assembly line and scale production and bring about the age of oil and mass production. As soon as that constraint is gone, the same thing will happen.


And now that main constraint of building software of coding is gone. And I think we're now in the age of AI and in this period of, of massive frenzy. So the key thing is now that software is, has become much more scalable and much cheaper to produce, how do we, how do we approach this? How do we make sure that the organizations that get built up in a frenzy include your organization as well? Because what happened back then in the early 19, 19 hundreds, uh, is there were over 300 car startups in Detroit, right? And only some of them survived that frenzy and went past the turning point. So, and of course for me, like for everyone else, this, this is a early days in terms of understanding about this, but I've really looked at it in terms of three different productivity multipliers that we're gonna get from generic AI and large language models.


There's, they're the code agents, uh, that are gonna multiply how quickly we can deliver code, how quickly we can refactor it, how quickly we can modernize things. So really down at the code level, they're the value stream agents, and I think think of agents as connecting GPT and large language models to various tools that can produce outcomes. Value stream agents will fundamentally change how we manage at the team of teams level. And then strategic agents, you've heard Gene and Steve Spear talk about, uh, rewiring organizations. We're gonna be able to, of course, have agents that take what good looks like, such as in their upcoming book and actually implement that because apparently without ai we've been very slow to do it for most organizations. But the key thing is, I think what we'll be doing is not, is leveraging the machines, leveraging the ai. So at each one of these levels, it's not that we'll have a 10th of the number of programmers.


As Samm Altman himself said, he thinks that we're gonna have 10 times the number of, number of software with the same number of programmers, because of course, the, the appetite for productivity and for delivering great experiences, uh, through software, through digital is, is just massive. Uh, and I think the interesting thing is we'll have this at every level. So the question is, how do we not just get that leverage that we can from people being amplified through ai, through machines at the code level, but at the value stream level, at the leadership level, all the way to the most senior levels in the organization. So I'll quickly dive into, it's kinda my, my current work in progress on each of these layers. So in terms of the first level, the code agents, I think code generations here is here now. And if you're not leveraging it, things like GitHub co-pilot in your organiz in your organization today, I leverage you to do that and to, and to make sure your leadership removes the impediments on that, because this is how programming will happen.


We have a new breed of co-ops and, uh, that just came into our company to plan Planview. And they actually, in the selection process, they asked us, are we gonna be able to use a, a co-pilot GitHub co-pilot? And if not, then we don't want to come here. And that's how profound this thing is. Uh, the code brushes, you can already make your code much better. Code brushes will make it possible to do these things that are very painful, such as have a bunch of colleagues who've, you know, are still behind on their angular to, uh, react migrations. Those kinds of things will become automatic. All of this will be automated through these large language models, uh, prompts, as we heard Patrick debut mention today, prompts will become the code. So a lot of what we're executing, for example, uh, my, I had my data science team over the summer implemented sentiment analysis using the AWS sentiment service.


Now that's just a little bit of prompt engineering, not a summer's worth of, of coding activity, uh, and curating the data. So these prompts will become a core part of the way that we code and actually the large language models will just become a part of our programming environment. Uh, and, and, and really the way to dig into this, understand this better to, to dig into what auto GPT does when it's not just a single prompt, but you actually create a loop and a feedback loop with a large language model that's going towards some goal. So that goal might be, let's say, refactoring your whole application. And, or maybe, I guess one of the harder things you could imagine is, is doing a your, uh, your, your SAP system modernization. I think that would probably be the limit, but, but it is theoretically achievable. So I, I think this will fundamentally change how we write code and how basically the effectiveness with, with which we can do this.


Now above the code level is what's really interesting to me, and this is where most of my time is going right now, where we have the architecture, the organizational design, and the value streams. So really think of the team of teams level. So flow optimizations, the things that I care deeply about through value stream management and the, and the flow framework. What we've seen is that we're surfacing these for organizations and showing organizations you've got too much work in progress. You've got too many dependencies between value streams, you've got all this waste in the system, all these cues, but not enough is happening about that. So what we actually will do is now be able to automate flow optimizations to automatically reject work when the value stream's overloaded, to accelerate flow, to remove burden from the system, uh, to make sure that we've got investments happening in architecture and tech debt, uh, that perhaps the business is not prioritizing by making it clear to the business what will happen if you don't do that, but find that that feedback in a, in a very natural language way, uh, roadmaps can be automatically managed in this way.


So a lot of companies already know what good looks like in terms of striking that balance between tech debt and feature investment, risk reduction and so on. Uh, but now all of this can be automated by bringing those best practices from tech giants, from startups to every organization through automatically managing roadmaps through these kinds of copilots that we will also have for the code level. And then this next one just blows my mind every time I try to think about it. But with auto GPT, you can actually iterate and bring the copilot into the loop and have it managed to some goal. That goal can be making a piece of code beautiful or an architecture better. We'll be able to do that at the value stream level itself. So set an objective and key result of saying we want to increase retention, uh, by driving adoption of this particular product.


Uh, so let's just automatically generate the features that will do that to that goal iteratively, until we see that and have, create that fully closed feedback loop, uh, in terms of delivering value to a customer while having basic people directing that rather than implementing it at, at a more detailed level. And those that is actually when value streams themselves will become generative. So the productivity increases from this are, I think are truly mind blowing. And then of course, why stop there, because we can bring this right up to the organization level where the agents are actually, uh, uh, the agents are actually ITing on the organizational design and structure itself. So today we do reorgs basically with, you know, through PowerPoint. All of this data is available, all of the flow metrics can be, are, are visible to an agent, and they can actually help refactor the organization to, let's say, target a goal of increasing happiness.


I know whenever we do any restructuring, any team changes, we always look at the employee net promoter scores for every value stream and target to that goal. But from having done that for years now, I know how much more effective a large language model would be added with, of course, the guidance of, of the humans behind it. Um, technology optimization, of course, you know, the whole notion of tech debt is, is very different when you can actually modernize applications, your whole technology stack, allowing your organizational design and your software architecture automatically. Uh, and another fascinating one is, is investment allocation. Too often the way that budgeting is done is without any real view of capacity of flow. Of course, all of this becomes connected. So that budget allocation, capital allocation in the organization is also supported by copilots that have much more depth and much more best practices, uh, and much more visibility to all of the data than than any single human can have.


Now, the challenge is that most organizations are just not ready to scale this. So this is real data that we see of value streams using our tools. Uh, and what we're seeing across our customer base is that 8% of the green box there of the time is spent in develop, of the time is spent in development. So if you now 10 x basically your development capacity, you're not speeding up at all. Uh, you've got these large upstream and downstream bottlenecks in terms of approvals, processes, you know, waterfall compliance and such. So the challenge is most organizations are, are not set up to do this. And one of my main, I think, calls to action is we need to move faster on this front or more organizations will get left, will get left behind. We've got more detailed data now. So with the five year anniversary of project to product, uh, we actually ran, uh, uh, a study combining both data from 3,600 value streams across 34 organizations.


So it's all live system data from Jira, ServiceNow, GitLab, GitHub, all those sort of tools as well as survey data. Uh, and we put those together. And what we noticed is that there's a 10 x mismatch between the ma the demand put on technology teams and the capacity of those teams. Because of course, whenever everybody plans, they don't take into and creates their strategic plans for the year. They don't take into, uh, account the capacity. 40% of team efforts are wasted due to overload and bottlenecks actually largely caused by that mismatch. Uh, too much work in progress and too many dependencies. This last one's shocking that even though we're over a decade into DevOps best practices, 30 to 40%, uh, of that time, that flow time's actually spent post, post, uh, code being written and deployed, it's usually just getting deployed into staging or development and actually not get into production is what's hap what we've noticed happening.


80% of value streams don't proactively invest in tech debt. And again, this is a place where we can leverage the computers, leverage the machines to make that, that that non-negotiable to reduce that, that tech debt and to always invest in it, uh, and to remove the burden from the system. And 90% of value streams measure outputs instead of outcomes, which of course is a challenge because it's, we need to bring those outcomes to the teams to know what they're contributing to, who the customer is. Uh, start with a customer in mind. And we can't leverage the benefits of value stream level agents unless we actually have that outcome there, have that outcome connected. So the challenge is right now is we've got those, if your organization has these three layers disconnected, that's a strategic layer, that value stream layer, uh, and that code team layer, you're not set up properly to leverage everything that's coming, everything that's there already.


'cause I think at the bottom, the code copilots are there already, and what's coming is these value stream level agents and these strategic agents. So I think the main thing is make sure you connect these layers. Make sure you've got OKRs flowing down. You've got visibility coming up across your entire tool set, um, to leverage that 10 x productivity gain that's there at the code level. Now, if I think, again, at that value stream level, I think there's another, another 10 X to be gained. And at the strategic level where the organization's on that being rewired to drive better happiness, better outcomes, better delivery, another 10 x. So, uh, those things will multiply in some way. That's, that's just, that's very mind blowing to me. But I think the key thing is set yourself up for this because all this is coming soon. So in terms of guidance and, uh, in this case, in the end, this is about people.


So this is, uh, a Vincent Van Ho Van Hook representation of, uh, Maya LeMans talk this morning, uh, where you've got a pizza party. So make sure that you do actually bring the people together to understand this, to plan it out, create your AI committee yet, and really focus it on each of those level, create each of those levels, both that, that code level, that value stream level, and the the strategic level, uh, and make sure that people are learning and that people are adapting to this new way. Because just like no developer is going to be successful five to 10 years from now, uh, without, uh, without leveraging ai, uh, leaders, technology leaders won't be either. And I think this will really will elevate what we can do, the value we can deliver to our customers and to our organizations. So to do that, to set yourself up for this, you have to connect all of the work from strategy to delivery.


That data has to be available, it has to be connected, and it has to be able to flow to a platform that can leverage it. Uh, you to do this, you also need to complete the shift from project to product. If you don't have product value streams, if you're still working in that old way of throwing work over the fence to, to it, it won't work. There won't be anything to optimize is the challenge. And, uh, I would, again, just basically as soon as you're, you get back home, I would just make sure that your organization is adapting what's there already. What today, the code agents like GitHub copilot are there to be today. And it'll help you learn across the organization, across your teams how these things work and how much you can leverage them. Make sure to, as you're doing this, to measure flow, to measure the productivity so we have a sense of the baseline that's there and how much productivity is there to be gained.


Because in the end, we need to make the economic case to organizations, to, to our leadership in our organizations, our boards on how much this can help us accelerate and keep in mind that others are doing this already, right? So in that survey, what we noticed is that a third of organizations that were there have actually made the shift and are in various levels of scaling from project to product. So it's happening. The value streams are there to be optimized. It's the two thirds I'm most concerned with because again, the guidance in the seniors has been there. It's there in terms of adopting best practice on agility, um, uh, and on DevOps and everything that's to come with generative ai. And LLMs is going to build on top of that. So to learn more, you can check out, uh, flow We're gonna have where in terms of how you can measure and create and set up those value streams.


And, uh, we'll be now creating some, uh, posts and announcements on the plan view blog as well in terms of those value streams and strategic levels and how you'll be lever able to leverage that within your own organization. So I think it's very exciting times and uh, you know, I know some of it can seem a little bit scary because again, I've never been through this much change. I have never watched this many YouTube videos in a, in a two month period before I used, I was actually trying to avoid watching them as much. But I think it's, it's more exciting than scary, especially if we can make sure that your organizations and the broader economies leveraging all of these developments, uh, to, to build truly great things. So with that, thank you very much.