Simone Cicero

Hello everybody, and welcome back to the Boundaryless Conversations Podcast. On this podcast, we explore the future of business models, organisations, markets and society in our rapidly changing world. I’m joined today by my usual co-host, Shruhi Prakash. Hello Shruthi.

 

Shruthi Prakash 

Hello everybody.

 

Simone Cicero 

Great to have you here. And our guest for this episode is Lee Bryant. Lee is a strategist, an organisational designer and developer, and the co-founder of PostShift. Lee is widely recognised for his pioneering work at the intersection between technology, networks, and organisations. 

 

And we have known each other for several years, maybe more than a decade. I used to invite Lee to the conferences we used to organise. So Lee, it’s a great opportunity to have you on the podcast.

 

Lee Bryant 

My pleasure. Really great to see you both and looking forward to a conversation.

 

Simone Cicero 

Thank you so much. At PostShift and in general in your career, you have been helping organisations to move beyond industrial age hierarchies towards more adaptive networked digitally enabled models. So you have a very long experience now. And in the last few months, we have been following your work, and how you try to integrate through your writings, and also the work you do with customers like we do, you know, this crazy important technology like generative AI. 

 

So I think we would like to start giving you a space to tell us a little bit more about what your experience has been, you know, based on your multi-decade-long work experience as an organisational developer at the edge, let’s say, of the space. 

 

So what are your learnings, and how are you now looking into the opportunities and the challenges that are opening up in front of us as we rethink organisations for this age of AI and beyond?

 

Lee Bryant

That’s a great, very big question. I think the probably useful context is that the way I came to this is a little bit different. So, although I’ve been playing with computers since the age of 10, which is a very long time ago at this point, and programming and so on. Actually, I was involved in politics. My first career was in politics and diplomacy. One of the things that really fascinated me there was the different organisational templates – from bureaucratic international organisations to very hierarchical government organisations to what we’re seeing today in Ukraine, actually, which is a more distributed networked, agile approach to empowerment and self-representation in the face of big, powerful bureaucratic forces on the other side.

 

So I was doing this work – it was crazy work. It was very intense. And I came across the internet, in about 1994, I think, when I was doing this work. And I started using that to distribute sort of blog-like information to all of our journalists, politicians around the world who were involved in supporting what we were doing. And it was immediately clear that this was a game-changer in terms of giving people the power of self-representation. So our information didn’t need to go through the prism or the filter of, you know, the BBC or you know, somebody that would try and sort of in a way compress the information to an acceptable form, we can actually speak to people directly and we can build networks of common purpose and get things done. 

 

And we really got things done. You know, we got things done with very few resources that were, you know, defeating more sort of powerful opponents in a sense. And so when I, I sort of stopped that work initially myself and Olivia, my business partner, we were interested in working with developmental organisations, with charities and so on. 

 

And instead of them being sort of basically direct marketing companies, that do some good work, we wanted them to actually connect directly with people, capture that purpose, that energy and do things in a more authentic way. Quietly, we drifted into the world of business because we also realised that there was a sort of humanitarian question in all of these sort of battery chickens sitting at their cubicles having rather sad lives and then going home and, you know, having a gin and tonic and a cry, you know, at the end of an awful work day. So we became very interested in actually how we could improve the experience of work. And I think one thing I recognised was that from a historical perspective, the late 20th century, or really, actually, the late 19th century sort of factory model really came out of feudalism in a sense. 

 

This idea that there are, you know, lords and peasants, there are sort of managers and workers, there are different classes of individuals, and the role of the manager was to recruit all of these workers, extract value from them, and make profits and so on. And it just occurred to me that, you know, with technology, and also with the rise of, education and the fact that we’re not just factory workers anymore, we’re knowledge workers to a large extent – we could do a lot better than this. 

 

So from the late 1990s onwards, we started building web-based collaboration systems, knowledge sharing platforms, ways of bringing networks and connections to the work of organisations. And we were quite successful in doing that. I think we were quite early in doing that. And then we saw the rise of the first blogs, and we realised that that was another game changer. 

 

You know, this was a almost free in monetary terms, form of networking and communication that was largely open source and it did the things that we were doing for lots of money. It did it really easily and really simply. So we switched, we created a new company called HeadShift, and we were really focused on injecting social technology into the enterprise as a way of humanising the enterprise, but also as a way of reducing its cost base, making it more effective.

 

And sort of in a way, taking back a little bit of power from this amorphous sort of class of generic managers, who seem to hold, all of the control and all of the power, but in fact, were not technology aware and therefore were not actually doing a great job in terms of the transition, to the internet age and so on. 

 

That’s really where we began. We had a great time with HeadShift. The company was acquired by a bigger US group. And then we sat and reflected and thought, well, what we’ve done is we’ve used technology as a kind of Trojan horse to create these outcomes that we were looking for in terms of more human connection, more agile working, et cetera. But where are the templates? Where are the organisational models that can be like operating systems in the internet age?

 

And we did some reading, we looked around, there wasn’t a great deal. There were some very old, excellent books from cybernetics in the early part of the 20th century, to Peter Senge, all these sorts of things. But actually, the only way that the most visionary organisations were able to make this work was if they had a visionary CEO. So all of the case studies, whether it’s Morningstar, all of these sorts of things, they really relied upon one extraordinary individual giving power or sharing power. 

 

And we thought, well, that’s an edge case. How can we make this more popular? How can we make this easier to achieve? And so we became interested in organisational transformation, templates, models, and so on. And then we did a lot of work with some very big companies, helping them with that transition. I think the way they understood it to a large extent was about agility. But actually, our goals were a lot wider than agility. They were about a sort of human purpose, about being technology-led as well as just being more flexible and adaptive. 

 

And I think what was interesting was, you know, we felt we had a very good handle on how to do digital transformation at a fairly large scale. But what we’d really underestimated was the inertia and the resistance of, management as a group within these companies, taking a, often quite short-termist attitude, thinking about themselves rather than thinking about necessarily the future of the organisation. 

 

So that was our, I guess, our blocker to having a greater impact in that area. And then one of the concepts that I know we all share is this notion of the platform organisation as a starting point, as a basic template. And what I like about the platform organisation is that we get machines to do the machine stuff so that people can be free to do the stuff that they are good at. 

 

Whereas in the old model, we treated people as fungible resource workers, right? They’re just following a process. They’re not really having much input. They’re not using their brains or their creativity. So essentially, we built a machine out of sort of human meat. And what I want to see is us using technology to build a machine upon which humans can be human connected, express themselves, and do the brilliant work that we all know they’re capable of. 

 

So that’s what I like about the platform concept – it’s the separation between the service layer, the automation underneath, and the freeform, agile, connected lateral networks of people on top. 

 

And when we were, towards the end of our period with Headshift and Datris Group, a friend of mine, Dave Gray, who was a colleague, we were working together. He wrote this book about the platform organisation, which I found really, really interesting. Very simple book, but it was a really excellent observation. So that became something we championed, from that point forward. But where AI comes into the picture is that the things that we were trying to get companies to do manually, which is map your processes, map your service layer, map your capabilities, try and automate the services underneath so that you can be less bureaucratic on top. All of that suddenly became a lot more achievable because the whole idea of agentic AI is that we can build little, very simple sort of bots or software workers that can own a service, oversee it, improve it, connect it with other services, and then we can actually achieve that service platform layer much more easily than we could in the past where it relied on lots and lots of change work, lots of transformation work, lots of consulting, lots of technical development and so on. 

 

So for me, I’m sort of at the moment in two minds. On the one hand, I can see the incredible potential of agentic AI to create the platform organisation and therefore to really become a an organisational operating system that others can adapt and they can use for themselves. That’s my optimistic side. My less optimistic side says, you know what, maybe companies are like nurseries or high school. They’re fundamentally a social competition structure. They’re not really about value. They’re not really about work. They’re about people feeling good about themselves. And that’s why they’re dominated by this class of managers, in which case, I’m just going to give it up and go home and do some gardening. So I guess that’s where I find myself at this point.

 

Simone Cicero 

Right. I mean, that’s a lot of stuff. So let me try to maybe ask you a couple of clarifications. 

 

So first thing, I noted in my notebook, you explicitly saying that agentic technologies, which it’s also important to underline that you’re not talking about generative AI in general, but we’re talking about agents. And you speak a lot about agents ops on your newsletter. You speak about agent-human interaction. 

 

So I think that’s a nuance that is worth maybe deep dive into. But I’m curious to hear most practically how you see agents as a technology that can extend the concept of the platform organisation that you have introduced. 

 

Because I’ve loved that you introduced this idea of platform organisation as the place where machines can do machine work and humans can do human work, which is very resonant with our framing and understanding of platform organisations being an entrepreneurial space, right? 

 

So really enabling the other people to do the work. So I’m curious to hear, first of all, how do you see agents, which are quite radically new technology, enter this space? Because they are a bit of a mix between machine and people, to some extent, if we look into them from maybe a naive perspective. 

 

So what do you think about it?

 

Lee Bryant 

Yeah.So I think if you look at, there’s a lot of really interesting things about AI and the idea of the programmable organisation is one that I will probably talk about a little bit more in this conversation because I think it’s a fascinating idea. But if you look at the evolution of technology, I think there are lots of lessons in the evolution of the software industry that organisations can learn from and probably will replicate over a period of time. 

 

One of those is the idea of microservices versus monoliths. So in technology, we used to build these big single platforms where everything is very closely coupled together. Then if there was a problem or a strategy change or you need something new, it’s actually quite hard to adapt a monolith. And so microservice architecture became a very popular thing.

 

I don’t know when, 10 years ago, 15 years ago, I don’t know. And the idea is that you try and build the smallest possible unit that can actually run a service. So think about being a finance company. Think about being a law firm or an accounting firm. You have a bunch of calculators, something that calculates utilisation, for example.

 

Now in reality, in an accounting firm, there are probably 25 versions of this. They’re all in spreadsheets and they live on someone’s PC, right? What you need is one service, which is a utilization service where you plug in some data via an API, and it will send you back a calculation.

 

And if you need that utilisation calculator along with 10 or 15 other calculators to build something bigger, you just send calls to the different services and you get back what you need. It’s an issue of composability, the idea that we want building blocks on which we can innovate and build new things easily without having to build everything from the ground up as we used to do.

 

So if you have in mind the idea of microservice architectures, I think this is the link with agentic AI. You know, an LLM is a model. They’re very powerful. People claim they’re going towards some kind of general human intelligence, which I actually don’t believe. But if you take a small language model, something that’s trained on accounting information or trained on legal information, then it’s a lot more reliable. It’s a lot smaller. It’s a lot simpler.

 

And you can create an agent and you tell the agent, your job is to run the utilization calculator. So I want you to be in charge of it. I want you to think about improving it. I want you to spot bugs in it. And I want you to make sure it’s, you know, five nines available to the rest of the platform when they want to use it. And all the agent does is it, owns that service and it is in effect that service. Now that’s a simple one. 

 

You can also have agentic capabilities, which are multi-agent. You might have one agent that’s an orchestrator, one agent that’s a planner, and then you have a series of specialised agents that do things that they are trained on and that they’re good at, basically. 

 

So if we go back to the platform organisation, one of the techniques that we’ve used very heavily in organisational transformation is capability mapping. We look at an organisation’s strategy. We say, what does the future look like? What do you need to do? What are the gaps in your capabilities? 

 

And typically maybe they’ve outsourced some stuff that they really shouldn’t have outsourced because it’s now core competence, or they buy in stuff. And we say, well, look, you need to own your core capabilities. You need to unpack them and understand what processes, what skills, what elements or components make up those capabilities. And then you have a working, in a way, process map of the organisation if you put that together.

 

But the beauty of agentic AI is you don’t need to do that manually. You can have agents that actually sort of represent these capabilities and make them available to other agents, to other systems, and of course, available to people as well. So that’s the way I see the advance of the platform organisation is that we effectively build an agentic sort of layer that does all of the boring process work that people are currently employed to do. 

 

Now, then some people say, well, what about jobs? And then the answer to that is, well, really? Do we want our ability to eat to be predicated on a nine to five job in a cubicle that involves two hours of train travel in a city like Rome or London or something like that? No, I think the whole world and the whole economy is changing. And we will not be solely reliant on these jobs in order to eat and survive. And anyway, these jobs don’t pay enough to buy a house, right? Or even to buy an apartment in many modern economies. So that model is anyway breaking. And so I’d rather see a world where we use agents to free people up and do better work. 

 

Maybe there are fewer people, you know, but they have a much more rewarding and purposeful job. And then other people can focus on their side hustles or their drop shipping or their entrepreneurial activities to try and support themselves. 

 

So yeah, so I’m quite positive about agentic AI. I’m a lot less positive about the impact of generative AI in the consumer space, but I guess that’s a different question altogether.

 

Simone Cicero 

Right, this is the week of the slope, so that’s a good point. But it’s maybe going to be not very in line with the releases when we publish the podcast. 

 

Coming back to the original topic that we discussed, you don’t mind, I have another quick consideration to share. So what I understand from what you say is that you see agents becoming the parallel to microservices in the age of microservices. We can call it micro capability providers.

 

So you increasingly look at the organisation as a set of capabilities, which is essentially running a process or processing certain information to certain outcomes, especially knowledge-related context, using a certain infrastructure in a much more heavily dependent context or machinery or things like that. 

 

But essentially, there is a capability that the organisation has that can be programmed because it’s a replicable type of job or it’s a reasonably clear standard procedure that you can put in the hands of an agent. So you can have this capability, suddenly be able to run 24/7, servicing everybody or supporting anybody that needs it. So, looking at the organisation first as a set of modular, composable capabilities. 

 

So my question – or the reflection rather that they have is what happens if we look at the organisation as a set of capabilities from two perspectives? First of all, it makes me think that the human contribution is gonna be increasingly about how you compose. It’s rather stopping about something being about doing the work, but rather choosing what work and how to compose it together to generate new outcomes. So it’s much more about composing. 

 

And when we look at our contribution as humans in organisations, as a composition, instruction, providers, and so on, it really becomes hard for me to think about organisation in the same way. So especially from a perspective of the boundary of the organisation, because first of all, I think that when I need a Lego brick, in the composition and building, why just using the ones that I have in my box? 

 

I should maybe look into generally what’s available on the market and other boxes. So it’s beyond my organisation. So suddenly, the work is no more about organisation, but rather it is about how do you compose across organisations? And secondly, what happens when the composer is also an agent? How do you see that dynamic really upholding the concept of an employee and the concept of an organisation?

 

Lee Bryant 

Yeah, it’s interesting because I don’t really like the concept of the employee. I like the concept of people very much, but I don’t like the idea of a role in a title which absolutely has baked into it a kind of inequality, right? Or boundaries on what they can achieve and what they can do.

 

Several things. I’m to mention a few things that come to mind there, perhaps without structure. One is that the majority of the most powerful innovation these days is combinatorial innovation. It’s precisely combining, remixing, composing. That’s what’s cool. You can look at the Gutenberg press, can look at the internal combustion engine at the end of the 19th century, and you can look at the iPhone.

 

In each case, the basic components existed, but it was put together to deliver a new customer experience or a new user experience with immense imagination and design talent. All of them changed the world, right? All three of those inventions completely changed the world. So I’m a big believer in combinatorial innovation, but to do combinatorial innovation, you need the building blocks and you need the freedom and the time and the permission, if you like, to think about those things. 

 

So in the platform organisation, I’ve always imagined that there are lots of people who are, you know, to take the sort of Haier or the Rendanheyi approach, they’re very zero distance to the customer. They’re basically located with users or customers, and they are really focused on those needs. 

 

And then what they do is they pull from the platform, components, agents, capabilities to actually create something entirely new and bespoke that meets a particular customer need. I think that’s a noble activity. I think it’s a great way of seeing the role of people in helping customers, solving problems in the world, and creating things that people use. The stuff at the center, the admin stuff, is sort of in a way less value creative. It’s less important.

 

So that’s the first thing to say. So yeah, there may be fewer sort of employee drones, just clicking buttons and filling in spreadsheets, but there will be lawyers, there will be accountants, there will be designers, there will be, you know, composers of all of this beauty and all of this wonder. And so I think that’s, you know, still a very positive, you know, vision.

 

I think it actually does relate to the studies that you’ve done on Rendanheyi model. It’s a great model, right? It’s got lots of natural components to it. It’s got lots of dynamics that seem to work very well. And so I think it makes that vision, as well as just a general platform organisation vision, a lot more achievable. I’m quite into that. 

 

But I want to make another point, actually, which may be a bit vague and philosophical, but it’s about how we build on things. If you look at human language, it began as a series of sounds and noises. It was then about, accounting largely what people were complaining about not having 10 bushels of corn when they paid for 12 bushels of corn or whatever. And what’s happened is, you know, you then abstract the language further and further until you get to things like poetry. And a poem is a bit like a zip file, right? It’s a form of compression. In a hundred words, you can evoke a thousand words because we all know what the words mean. 

 

They refer to existing knowledge or cognitive services in our minds. And so it’s a beautiful way of building on the shoulders of giants. And this is also what software has done. You know, software started when I was programming, it was machine language, it was basic, you know, then it was all of these low level languages. Then we have frameworks like Ruby on Rails, you know, and now it’s, you just type in NPM or whatever into a command line and it will build you an entire infrastructure, but it’s still a technical domain. 

 

And now we’ve gone the extra step where we can do all of that using natural human language. And so in a way, we’re moving up the abstraction layers, we’re moving up the value chain, and we’re creating the ability to do incredible things with small instructions because you’ve got all of these layers below you. 

 

And so if you imagine what a team of 10 people can do, if they are able to instruct their big orchestrator agents, using natural language and using their own culture, their own values, their own strategic goals, and then below them can unfold this complex, detail of how that stuff happens, how the app gets built or how the business unit gets built or how they do the market, all of that stuff. 

 

These are superpowers, and so we need to make sure we use them well. And that’s why I’m also interested in sort of enterprise AI. It may sound boring compared to the wider world of the internet, but in organisations, you know, people have a common purpose, they have mutual responsibilities. They have contracts. You know, that’s a very, a very higher concept, the idea of mutual accountability through, through contracts and so on. 

 

So organisations have this ability to write their own constitutions, their own code, their own rules, their own cultural norms and values in a way that can be good and it can be bad, but at least it’s not the open wilds of the internet where consumer AI will be used to trick people, scam people, overturn governments, create populist movements, all of this negative stuff that we’ve seen with the rise of social media. 

 

So this is the other bit I’m actually exploring right now is the beauty of rules and the beauty of actually making the implicit explicit, right? Because if you make your norms and your rules explicit, which by the way, you need to do for agents to be able to operate because they don’t assume anything, then actually you also create a more equitable environment of work. This is the one positives that Max Weber, said about bureaucracy was in fact, people know where they stand, right? There are some rules that are supposed to be enforced and actually you can rely on them rather than the powerful whims of the leaders, or the warlords, or whatever we’re talking about at different times in history. 

 

So I’m really interested about this intersection between language as programming, the explicit nature of rules and norms, and then what we can do with all of that in order to make work easier and to do more value creation rather than sort of performative workplace theater, turning up to meetings, communicating in PowerPoints, and just trying to work out who’s more powerful than who.

 

Because none of that, it’s all socially interesting, but it’s not interesting in terms of value creation. So I know that covered a lot of ground, but those are my sort of interests.

 

Simone Cicero 

Yeah, but I think it’s extremely resonant with some of the things we released more recently. We wrote about more recently, you know, the idea of constraints, the idea of clarity of rules and instructions. So totally, mean, but I want to leave some space to Shruthi, who has some questions coming up. 

 

Shruthi Prakash

Yeah, I’ve written too many points now. So if I ask a question from before, don’t mind. yeah, mean, few things that stood out first, I wanted to tell right from the beginning what you spoke about power of, let’s say, self-representation, network of common purpose, and bringing in humanitarian side to the organisation. 

 

So that was on one side. And then there is the other side of, let’s say, using technology or AI in this case for very technical work, let’s say finance, tax models, or any microservice sort of architectures, right? So where do we balance the two? 

 

I think that’s something I wanted to touch upon that, you know, whether to look at AI from simply like automation point of view, which I know a lot of leaders do. So do we stop there or do we change that language to AI then augmenting human capability, redesigning you know, organisational workflows, structure, decision making, team dynamics, things like that. 

 

How do we sort of do the two and yeah, balance both of them? I think that’s something I’m curious in a very practical sense in terms of application at an organisational level.

 

Lee Bryant 

Yeah, that’s a great question. And that balance is going to be something that I think we all explore and that we will make some mistakes and we’ll go down some blind alleys along the way. 

 

But I think, first of all, we shouldn’t be afraid of automation of soulless, boring, repetitive tasks. I don’t think anyone should really be afraid of that. So the two things are happening in parallel in a sense. I think there is work going on. We’ve had RPA and other forms of automation for years in organisations, but I think now what we have is smart automation where you don’t just automate the service, but the service has some intelligence and self-awareness and goal-seeking behaviors because that’s intelligence devoid of goal seeking is not intelligence. It’s just pretend, right? 

 

So I think smart automation will continue. It’s a really important thing. It’s not hard to do. It’s a lot easier to do now. But the augmentation is the really interesting bit. And I think that the problem in the middle is how we’ve defined work. And because work is all about jobs, roles, and processes to a large extent, it’s fundamentally dehumanising from the beginning.

 

And so if we want to augment the human, we need to sort of in a way unlearn some of that mental model, right? We need to just say, look, this is our purpose. This is our strategy. This is what we’re trying to achieve. You know, go for it. Here are the rules. Here are the guidelines. Here’s roughly how we operate, but you know, go for it. And I think what we will see is everybody being a manager because everybody will manage agents.

 

They may manage people as well and there is some value in that for good managers for sure. But I think people will have a more empowered relationship with their tools compared to the first sort of two waves of enterprise technology, which were all about big platforms that IT spend, you know, five million euros building and it takes three years and it’s a load of rubbish, but we have to use it anyway because that’s what we’ve got, right?

 

So I think that’s one of the empowering ideas that we’re gonna see here is more human control over the technology because things are programmable. And I think that in itself changes the power dynamics, it changes the social dynamics, and therefore it has more potential to augment human work. But when we say augment human work, it’s a bit like the field of HR. 

 

You can’t empower people in the sense that you can’t do it to them. So I can’t as an enlightened manager augment somebody. They need to be fundamentally empowered in the first place to use whatever they can get their hands on, in order to elevate themselves and elevate their own work and their own creativity. So the language is always gonna be complicated, but I don’t think there’s a contradiction in the two. I think we’ll see a lot of basic boring work essentially eliminated through automation. 

 

And by the way, that includes probably 30%, 40%, 50 % of middle management as well. Because actually, if you write down an algorithm about what they do, it’s like, look at sales numbers. If number too low, have a meeting, shout at someone. Agents can do that pretty easily. So we’re going to see that level of replacement of tasks for sure. 

 

But as you hinted at, where it gets really interesting is where we’re going up the stack a little bit and people are learning how to use these capabilities to do things that we can’t even imagine, you know, or things that would have taken a whole department in the past in order to fulfill. So I see them both working hand in hand. think we’re to have a lot of tensions. There’s going to be a lot of misuse of some of these ideas, but I think the potential is there, and it’s really going to come down to natural competitive dynamics, you know, where certain new companies

 

certain transformed old companies actually demonstrate success with this and then I think it becomes a lot more mainstream at that point.

 

Shruthi Prakash 

I mean, I’m curious to see because even for me, the language we use in organisations just from a structural point of view, right? It’s such a unfortunate sort of command and control reporting lines, like your everyday language, just how you speak in organisations is still outdated. So when you put tech on top of that, which is so dynamic, it becomes harder to adapt. 

 

So I’m curious how that works.

 

Lee Bryant 

Well, this has been the problem, right? Because every wave of technology has been wrapped around the existing corporate hierarchy. You know, from IBM, you know, IBM got started really doing sort of calculations on missile systems and so on, and then the mainframes and then the sort of middle tier and then the PCs and then the internet and every stage has been wrapped around the old hierarchical structure. And that’s been the problem, right?

 

And that’s still a risk with AI today. We could just automate away some lower level jobs, but still have this sort of absurd sort of Heath Robinson contraption of a hierarchy on top that believes it’s doing work. So that’s the key thing is we need to reimagine and rethink from first principles. What is work? Why are we here? What are we doing? What’s the goal? And then try and apply it more rationally.

 

Simone Cicero 

Yeah, I mean, I was reflecting on this idea that somehow we are transitioning in a world where we’re going to see much less friction, much less, in terms of capital, for example, right? We see a lot more leverage. You spoke about agents and everybody becoming a manager.

 

And you said that this situation caters to an organisational context where, rather than managing people, you set the rules and you try to convince them basically to do things in your system, right? So as organisational developers nowadays, the challenge we have is more like, okay, how can I create a system where people want to come and create? 

 

Because there’s so much potential outside that – why someone should actually join an organisation, which is more or less reconnecting with the idea that I was bringing up before, that the idea of composability, of infinite composability, is inherently challenging to the very idea of what an organisation is and what it’s not. Because to be honest, at the end of the day, if we look at the theory of the firm that was premised on transaction costs, right? Once this transaction cost and leverage goes haywire, how do you rethink the organisation? 

 

I don’t want to make this conversation so fuzzy and theoretical, but maybe you can help me to put it into a more practical perspective for leaders. 

 

But it looks like we are due for a deep rewiring of the organisation, right? What really is an organisation? So how does it look like in your point of view and what are the practical or essential ideas that organisational leaders may consider to be pioneers in this space rather than followers?

 

Lee Bryant 

So I think the first thing I would say is that although, back to one of your points, the external boundaries of the organisation may become permeable, so they’re less like boxes and they’re more like this is our center, but actually we bleed out, we stretch out beyond the formal boundaries of the organisation. At the heart of that is the organisation’s platform, right? 

 

It’s the, in a sense, the machine and also the core people who put in place its values, its value chain, and its sort of view of the world. And I think if you are a successful organisation and you build a platform to borrow a phrase from Tim O’Reilly, you create more value than you consume, then you can invite other people, other companies to use that platform as well, right?

 

This is the sort of Amazon platform model. It’s a model that’s incredibly successful. You build more than you need. You have, let’s say, an e-commerce engine, a marketing engine. You’ve got all of these different pieces. But you’ve got little companies like the birds that clean the teeth of the crocodile who are working at a level of revenue that you’re not interested in. 

 

But in fact, they could rent space on your platform, or they could use your platform in a different way that’s not core to what you do. And therefore, you would increase the influence of your platform – you would become more central in the market as a player, but you still get to focus on what you want to do. 

 

So I think the focus on understanding your capability needs, mapping them, and focusing a nonstop sort of permanent process on capability development, for example, in the direction of agentic AI, but not only in that direction. There are other non-AI technologies that are still incredibly relevant.

 

I think that’s the route to strategic advantage in any kind of knowledge-based work or any customer-facing work. Now, it’s a bit different for pharmaceuticals, for manufacturing, for Siemens making gas turbines at vast scale and so on, but the principles still apply. Rather than think about a web of manual processes that needs to be held together by the enforcement of management, think about building the machine that creates the machines, if you like. 

 

And focus on that platform development so that you can then allow all of the talent in your organisation to come to the table, really, and to apply itself to what we can do with these capabilities and what we can do with that kind of platform. 

 

This is where I’m quite optimistic, because I do think that there’s so much investment and attention being applied to agentic AI, that it will actually become a subsidised technology, right? A bit like Uber, you know, where I live, Ubers are super cheap and pretty much Uber has been paying you, you know, every ride because they’re going for that classic, you know, scale play. And I think it’s a bit like with agentic AI, we’re going to have access to a ton of really useful tech that billions of excess capital have been put into creating.

 

And that gives us a chance to build out these sort of engines, these value creation engines and these platforms at the heart of an organisation. And then we’re free to organize using whatever socio-political, socio-technocratic model we want, right? Because we’ve solved the problem of hierarchy in a sense. If you look at hierarchy having three purposes in an organisation, one is status signaling, which it’s natural at – Dogs do it, humans do it, even cats do it. Then you’ve got hierarchy as a communication network, which it’s terrible at. It’s just about the worst way to communicate. And then you’ve got hierarchy as a way of holding together these manual processes. Well, the third one we don’t need, the second one we’ve already moved beyond, and we have enough social technology in the organisation to communicate effectively. 

 

Yeah, sure, people can wear badges and different colored clothes if they want to feel important. Who cares, right? What matters is that we actually do the work.

 

And that’s where I think agentic AI coming together with the platform philosophy is just the most hopeful moment, I think, in the last 20, 25 years of digital transformation to actually create real 21st century organisational operating systems.

 

Simone Cicero 

So I wanted to ask you a last maybe question. I don’t know, again, if Shruthi, maybe you have anything else you want to drop in, maybe after this or before. But essentially, I was trying to go back into some of the points that you made about how do you make this platform, so you mentioned essentially, to get people to invest, contribute, and participate to a particular organisation, what’s going to make a difference is your platform. 

 

How great your platform side of the organisation is, how much are you enabling, or how do you make the employees or the participants to your organisation able to create value. Of course, you also made this connection with the value chain and, in general, the concept of value. So I was curious to hear from you – If you look at, if you think about an organisation and you talk to leaders and how they can make a difference, what does it really mean to make a difference on the platform side of the organisation in terms of enabling services? 

 

And what is the role of value in that? So is there a dual system where maybe what you need to make clear is what values the organisation wants to achieve? What kind of enabling systems makes available to the people for them to really seek and produce this maybe shared concept of value, also including some aspect of ownership of this value. So how do you close the circle to really make a difference today in building the organisation?

 

Lee Bryant 

I mean, on a very simple practical level, you one of the things that we do and that we recommend, leaders do is to, in a way, take a lesson from, you know, DevOps in the technology world, these sort of unshaven people wearing, you know, old t-shirts in the basement, know, managers never really paid them a lot of attention, but actually they were among the most professional organized people in the whole organisation. 

 

Because if they had to do a task, three times, they would build a script to automate the task, and then they would build a script to manage their scripts, and then they would build a controller to maintain observability over that system. 

 

And that’s how, we went from having teams of sysadmins desperately trying to run networks to having small teams running vast scale networks because of automation, orchestration and composability.

 

Those are the skills that they developed out of necessity. So there are some lessons there, I think, for leaders. And if you see somebody running a process manually three times, ask them to design a magical system that runs it automatically.

 

Because most people understand the bit of the organisation that they’re in. They can tell you what capabilities they have. They can tell you what exception handling they need to get done to perform the work. They are your design partners, actually, for automating those services. And if you can give them some reassurance that they’re not immediately going to get fired, right? And that if they automate that stuff, that’s the bane of their life, they can then actually focus on creating real value on top – then I think we can engage people across the organisation in a sort of distributed fashion to come up with a concept, lots and lots of little concepts really about the agents and the service automation that could be part of the future platform. 

 

And then, other simple stuff, don’t allow people to go off and buy individual technology platforms that are not integrated. Make sure people refer to this capability map and build something that’s integrated and that speaks to the other systems. know, marketing usually go off and buy their system, HR go and buy their system, finance buy their system. 

 

So these are some practical steps, I think, that can get us towards where we’re going. But then in the, I do quite a bit of leadership executive education and, you know, mindset changes are a key part of that. You know, leaders as architects of the workplace, not just bosses. Leaders as navigators of uncertainty, not just people that try and predict the future and then plan or Gant chart their way towards towards meeting it. You know, these are sort of new leadership skills, which in a way are also old leadership skills. They’re quite traditional strengths of naturally good leaders, but we need to talk about them. We need to cultivate them and we need to move away from the concept of, just sort of bureaucratic shuffling and monitoring. 

 

For example, one of the lessons of open data is you don’t need enforcement if you put the data on platforms and on screens and on walls and everybody can see where things are going. If everybody can see where things are going, they tend to self-manage towards the right outcomes because we all know what we’re supposed to be doing. You don’t need quite so much enforcement as you would in a closed system. 

 

So I think there’s some very practical steps that can be taken by leaders to get this ball moving but they really do need to engage with the technical detail, not to the level of AI, MCP, know, RAG systems, but just conceptually. They need to understand, you know, how things work in the 21st century, even if they’d never touch a line of code. They need to have that confidence and so one of the things I do a lot is try and bridge that gap between the world of technology, the world of, let’s say, you know, collaboration and social organisation and the world of organisational design from a, you know, sort of straightforward commercial point of view, you know, what’s the value chain, what’s the investment model, what’s the ROI forecast and horizon scan, et cetera. 

 

So I think we need to bring those three things together. And, you know, there are remarkably few people who can actually work with confidence across those three areas. And I think that’s one of the problems we’ve had the vertically divided organisation for so long. We’ve had swim lanes, we’ve had over specialization for so long that building the laterally connected platform organisation is, you know, quite literally, you know, a 90 degree shift. And that’s that’s going to take some time to work through the system, I think. 

 

But there are lots of social factors that also feed into this. You know, the pandemic changed the world of work. People realized that if they didn’t go to work, the world wouldn’t collapse. Work from home and remote working is something that is now accepted as a normal way of working. And so many people, going back to your Ronald Coase, the argument about why you would work in a corporation, so many people are deciding they don’t want to. And so the attraction of these corporate roles is reducing all the time. 

 

People would rather be a digital nomad, you know, go and live in, you know, in Bali or Jakarta or Lisbon or wherever you want to live. And so I think there’s a number of other factors in the background quietly, you know, moving towards creating the conditions for that, for that kind of change. 

 

And I think the other part of this as well is, you know, we’re often critical of corporations, but actually corporations are generally quite good corporate citizens in some respects. You know, if you look at Unilever and its social purpose, you look at the sort of internal rules, you know, and diversity initiatives of many large companies, we’re in a weird position now where actually the companies themselves are in fact better custodians of 20th century democratic values than some of the governments who are challenging them. You see that obviously with Apple in the United States. And so in fact, you know, we may be getting together in these organisations partly to create a little oasis, right? 

 

A little world building – where we can have some values that we believe in, but also we have a commercial purpose where we all want to make money, you know, at the same time. I play far too many video games, and the idea of world building is such a fascinating concept when it comes to organisational design, because the best games, know, Cyberpunk 2077, Borderlands 4 that just came out, they build a world that has sort of aesthetic, it has a vibe, you know, it has some backstory.

 

It has something that really attracts you. It’s like a good movie would do the same or a good long-term series would do the same on Netflix. And that kind of world-building, I think, is a really powerful force for employee engagement and for getting people to join you and believe in you. 

 

The platform organisation is not a purely technical concept, right? mean, you look at Rendanheyi, you look at the higher model entrepreneurial ecosystems, mutual responsibility and accountability, freedom to innovate, all of this stuff. It’s quite human at the same time. It’s very competitive. And that’s one of the things I like about it because competition actually brings out the best in people. It’s what makes us innovate. So it’s not a comfortable, safe place for lazy people. But the potential is there for anyone who wants to value and do something extraordinary.

 

So, you know, I think there’s a lot of factors here and a lot of things for people to focus on. But going back to the leadership point, on the practical level, yes, we can do capability mapping and we can get people to sort of automate their own work to an extent. But on the big level, we need to do some world-building. We need a direction and we need a concept that can attract people in terms of a good place to work, a good place to invest in and a good organisation that you want to buy from and be connected with. 

 

And so that whole world building thing, I think is also part of AI. And in fact, on a technical level, it’s also part of AI because the next phase of AI is world knowledge. It needs to understand the physical realities of the world around it if it’s to do more than just knowledge, work with data and words. And so we’re seeing that with driverless cars. We’re going to see that with industrial applications and so on. 

 

And again, it’s all equally, equally fascinating.

 

Shruthi Prakash

For me, this is super interesting because I love the idea of a distributed designer almost inside and outside organisations. And yeah, the idea of world-building. I think when you spoke about just digital nomads as well, I’ve seen that it requires both an internal and external restructuring you as an employee and then your employer as an employer essentially. 

 

But it sort of happened. I think it’s a matter of time. I’m happy that, you know, we’re sort of working towards that. 

 

So towards the end of the podcast, we have a section called as the breadcrumbs, where we ask our guests to share anything that inspires them, maybe podcasts, movies, music, or anything you do in your routine, maybe that, you know, is different and somebody can learn from it.

 

Lee Bryant 

I, it’s difficult for me because I’m, I’m a terrible filter feeder. I’m like one of those big whales that just takes in a million plankton. And I read a lot of trash. I read everything actually from geology to popular culture, to video games and so on.

 

So I’m not a classicist in terms of my book list and my book recommendations, but there are some people doing great work, I guess, at the frontier of both the business side of AI and the technical side of AI. So I am a fan of Azeem Azhar, exponential view. I think he’s doing something really great. I think Simon Willison is also doing something great on the technical side, also Nathan Lambert and people like that.

 

So I think it’s good to read that even if you don’t understand all of it because it’s just a way of getting into the topic. think on the sort socio-technical cybernetics layer, which is also fascinating, people like Henry Farrell and Dan Davis are sort of continually writing about engaging topics related to that. 

 

And then there’s all kinds of other people. My friend Céline Schillinger is writing good stuff about engagement leadership, Zoe Scaman about strategy and AI. But, in terms of books, I’m far too all over the place to, you know, in a time of sort of increasing populism and the collapse of democracy, then Milan Kundera is a good read. You know, for example, how to live in that. Ivan Krastev, like in the Financial Times about sort of taking a sort of centralist European perspective on all the crazy stuff that’s happening. 

 

And of course, being in Portugal, Fernando Pessoa and his book of disquiet because he adopts multiple different fragmented identities to explore all of the dimensions of his soul and his personality. And that’s something I think we’re all experiencing right now. Seeing the different sides of how we think and how we work as person. 

 

But yeah, I have a lot of empathy for AIs. People criticize AIs as being next word predictors but I read so much trash that I am just basically a next word predictor. I’m just regurgitating popular culture from the last hundred years in everything I do so I can’t take any credit for my own ideas but I love sharing the other ones.

 

Shruthi Prakash

I also have never heard somebody describe a poem as a zip file and then relating that to technology, that’s very cool. It makes something very technical and hard to understand, easy to grasp. So really appreciate the feedback as well.

 

Simone Cicero

So thank you so much, Lee. It was really an amazing conversation. I think I resonated a lot with most of the things you said. I can see how our research agendas are really crossing to some extent. Especially, I want to double click the idea of organisational development as a world-building. 

 

I think this is a massively important point that you raised because it really conveys the idea that you have to connect as an organisational developer and designer these days. You have to connect both sides, the two major elements of the organisation, which is on one side, the technical, the automation, the building and the composing and so on that you mentioned when you said let’s look into what DevOps can tell us, for example, which I think it was a brilliant parallel. 

 

But at the same time, if I think about the people doing technical optimization, they typically work in a context where it’s very clear to them what is the outcome they want to generate. So for example, application availability or performance and so on. So if you don’t design – What we call the forcing function, so the constraints of the system, which actually translate your value system into what you want your organisation to optimize for, it’s really pointless to automate. So what are you automating for? Are you just automating yourself into a lack of strategy or a lack of value.

 

So I think it’s really clear that word building means both. It means creating the infrastructure, always improving the organisation. But in a context where you dictate, you design, even collectively, you design the value system. You design what you want to generate, what are the outcomes that you want to create, and you translate it into the constraints you give to the organisation. So I think that was a very important resonant point for me. 

 

So I hope you also, somehow share this perspective and liked the conversation and enjoy the conversation with us.

 

Lee Bryant 

Yeah, it was great. And the reason I enjoyed it as well is because I’m a big fan of what you guys have been doing. There’s real depth to that research agenda, and I’ve benefited from it, and I enjoy following it. So let’s both keep it up and stay in connection.

 

Simone Cicero 

Thank you so much. Thank you so much for saying that. It’s always great to know that we have been helpful. Again, Shruthi, thanks for your time. Thanks for joining today and for your questions as always.

 

Shruthi Prakash 

Thank you. Thank you, Lee. It was really nice connecting with you. Thanks, Simone.

 

Lee Bryant 

Yeah, and great to meet you.

 

Simone Cicero 

And our listeners, of course, you can check the web page, boundaryless.io/resources/podcast, where you will find the latest episode with Lee, with all the fantastic mentions and references that he shared during the conversation in the notes and the transcript.

 

And of course, until we speak again, remember to think Boundaryless.