Subscribe
Share
Share
Embed
As AI evolves from a productivity tool into an active participant in decision-making and execution, we explore what agentic organisations mean for leadership, organisational structures, and the future role of humans at machine speed.
This week, Dave and Rob sit down with Gidion Peters, founder & executive consultant Agentic Organizations and Transformation, to challenge conventional thinking about AI, asking whether our next colleague could be an agent, how organisations operate at machine speed, and what role humans will play in the agentic enterprise of the future.
This week, Dave and Rob sit down with Gidion Peters, founder & executive consultant Agentic Organizations and Transformation, to challenge conventional thinking about AI, asking whether our next colleague could be an agent, how organisations operate at machine speed, and what role humans will play in the agentic enterprise of the future.
TLDR
00:41 – Introduction
02:39 – Hang out: Rob is confused about AI copyright and agentic software development
03:49 – Dig in: What happens when AI agents become colleagues instead of tools?
07:40 – Conversation with Gidion Peters on agentic organizations, leadership, and the future of work
56:50 – Could AI agents soon have a place on your company's org chart?
Guest
Gidion Peters https://www.linkedin.com/in/gidionpeters/
Hosts
Dave Chapman: https://www.linkedin.com/in/chapmandr/
Rob Kernahan: https://www.linkedin.com/in/rob-kernahan/
Production
Marcel van der Burg: https://www.linkedin.com/in/marcel-vd-burg/
Dave Chapman: https://www.linkedin.com/in/chapmandr/
Sound
Ben Corbett: https://www.linkedin.com/in/ben-corbett-3b6a11135/
Louis Corbett: https://www.linkedin.com/in/louis-corbett-087250264/
'Realities Remixed' is an original podcast from Capgemini
Creators and Guests
Host
Dave Chapman
Chief Cloud Evangelist with nearly 30 years of global experience in strategic development, transformation, program delivery, and operations, I bring a wealth of expertise to the world of cloud innovation. In addition to my professional expertise, I’m the creator and main host of the Cloud Realities podcast, where we explore the transformative power of cloud technology.
Host
Rob Kernahan
VP Chief Architect for Cloud and Cloud Realities podcast host, drives digital transformation by combining deep technical expertise with exceptional client engagement. Passionate about high-performance cultures, he leverages cloud and modern operating models to create low-friction, high-velocity environments that fuel business growth and empower people to thrive.
Producer
Marcel van der Burg
VP Global Marketing and producer of the Cloud Realities podcast, is a strategic marketing leader with 33+ years of experience. He drives global cloud marketing strategies, leveraging creativity, multi-channel expertise, and problem-solving to deliver impactful business growth in complex environments.
What is Realities Remixed?
From the award‑winning team behind Cloud Realities, Realities Remixed explores what happens when people, culture, industry, and technology collide, because real breakthroughs rarely come from technology alone.
In each episode, hosts Dave Chapman and Rob Kernahan dig beneath the surface of innovation to understand what’s truly shifting. They speak with leaders, practitioners, vendors and thinkers across industries to explore how technology reshapes human behaviour, organisational culture, and society, for better and sometimes for worse.
Web - https://www.capgemini.com/insights/research-library/cloud-realities-podcast/
Email - realitiesremixed@capgemini.com
Is artificial and one of them is real, what would you what would you bet on?
Speaker 2:I'm gonna say the trailing flowers are artificial and the bouquet of flowers are real.
Speaker 3:That's what I said, but you're not you're wrong.
Speaker 2:Is it?
Speaker 3:You're you are wrong a 100%. This is a great start to the conversation, by the way, you know.
Speaker 2:I mean, mean, when you use that phrase, I'm wrong a 100% of the time. I mean, yeah. Okay. People have been telling me that for a long time. So
Speaker 1:I'm Dave Chapman. I'm Rob Kernahan. And this is Realities Remixed, an original podcast from Capgemini. And this week, we are deep in conversation about the realities of agents in organizations, the impacts they might have, and how we might have to change our organizations to get there. Now joining us later in the show is Gidion Peters.
Speaker 1:He is an independent consultant and expert in org transformation and specifically agendic organisations. Gidion has a background in running organisations and has got quite a lot to say about it. Super interesting. If you want to leap straight to our conversation with Gidion, as always, you can find the time code in the show notes. But before we get there, Robert's here.
Speaker 1:How are doing, Robert?
Speaker 2:Hello. Hello. I'm alright, actually. It's good. It's Friday.
Speaker 2:What's not to like?
Speaker 1:What is not to like exactly? Exactly. Any plans for the weekend?
Speaker 2:Taking it easy, the one task I have is my son is coming back for he's just been interrailing around Europe for two weeks. Amazing. He's been sending he's out he's good fun has been had with a group of seven friends, and I have to go and install the broadband at his house to configure the mesh network. So I'm I'm being tech support You configure the for myself, but that's it then and then I'm just gonna take it easy. How about yourself?
Speaker 2:Any plans?
Speaker 1:I'm definitely not configuring a mesh network. I mean, that I mean,
Speaker 2:that is the most fun you could have on the weekend. So what's one step beneath that?
Speaker 1:You've set a high bar with that. So the the one step beneath, I'm going to the Haywood gallery tomorrow to see the Anish That's Kapoor
Speaker 2:a bit eyebrow, Dave. What's happened there? Did somebody else get you tickets?
Speaker 1:I actually drove that one. A couple of times a year, you know. I tried to try to do something a little cultural. So I'm looking forward to quite a bit.
Speaker 2:Very good. Alright. Yeah. Yeah.
Speaker 1:Alright. Now, like configuring a mesh network, but what else is confusing you this week?
Speaker 2:Well, this is a bit of a complicated one, Dave, but it is an important one. So we've often talked about agent created or AI created media, music, film, videos, and there's a copyright concept and an issue and it's raging debate in society. But a new one is now lifting. It's a bit left field but people are starting to get concerned by it, which is we're moving into a world where a human describes an intent to an agent and the agent goes and writes the code. Mhmm.
Speaker 2:The problem we have is this creates a copyright issue because copyright protects human creative expression, not simply an idea or requirement. So if the human didn't meaningfully shape the resulting code, it is difficult to establish a human author and therefore difficult to claim copyright over software. Which actually means that when providers may rely on contracts, trade secrets and control, things like this, you're shifting protection away from public copyright law and towards private operational control. So a rather interesting debate has kicked off about if I used AgenTik software development for the majority of my thing that I create, I may not have protection and therefore I have to use different mechanisms to protect that. So it's quite a serious issue.
Speaker 1:Does it mean that in a future world someone might argue that the agent has copyright over that? Well, so this Is that where going with this?
Speaker 2:Yeah. This is the confusion part, which is when we bring agents into the system, are we going to have to update legislation to, you know, respond to the world that we face today? So copyright law's been around for ages and hasn't changed but now we have this agentic effort doing the majority of the process and we're going to have to update these things or do we just leave it and say, no no, we just continue to protect human creative expression. If you use an agent, well, so
Speaker 1:be it. Yeah. It could be as simple as, know, you use the tool to do it and you know, the tool might go one one or two steps off, 10 steps further than a traditional tool might have, but it's still the human orchestrating the tool. Or you could go to the other end of the, you know, the agent has some agency, no no pun intended, but then that introduces a whole other slew of things, doesn't it? Because that's where you get taxation arguments and various other things.
Speaker 2:Yeah. And and at the moment, copyright very particularly says can't protect the idea or the function, you protect the creative work of the human that resulted in it. So actually it's gonna have to be an update or if you don't, various things are gonna have to shift in the way we protect and control software. So yes, we've had the debate about creative media and music and the copying and the training and everything else, but actually they've picked up the software point now and gone, we might have an issue if you create a product using software, you might not be protected in the future. Which is becoming quite an interesting thought about then what people can do and get around it, etcetera.
Speaker 1:And and of course that would that would apply presumably to things like the arts. So we talked a fair bit on the show back and forth about the pluses and minuses of say AI created music or film. In that situation, do you have a similar copyright challenge, do you think? Who owns a copyright on a piece of AI created music?
Speaker 2:And this is the quagmire we're wading into that says we've got we've got legislation that's been around for decades now incompatible with the way the world might create output. And so whether it's software, whether it's creative media, you know, music, we're going to have to update things. Now, the problem with that is the legislation is always lagging and this feels like very complicated legislation to be updated. So what happens in the intervening period whilst the legislators try and work out what the answer to this is? And you might get this disruptive entity in the middle that might cause a load of major issues because we just haven't got the rules and regs properly in place.
Speaker 2:It goes to the wider debate about AI and the future of it and the human impact. But this is quite a serious side effect that's being picked up with legislation.
Speaker 1:Well, let's dig deeper into that type of subject matter with Gidion, where we're going to explore with him organisational ramifications of bringing agents in, kind of beyond the beyond the technical challenges of automation and, you know, kind of changing process and those things, but what, you know, what does it mean and feel like for for agents to be in organisations? So with that, let's go to our conversation with Gidion Peters. So Gidion, give me a sense before we get going of here on the production sheet, I've got expert in organizational transformation and agentic organizations. Are you an independent? Do you work somewhere?
Speaker 1:Have you got your own company? What what how does that show itself?
Speaker 3:Yeah. So I've been an organizational consultant for, for quite a number of years now. I used to run companies based on this. So I I had two of my own consulting companies, based in The Netherlands, the city of Utrecht, and and in The US. But I felt I wanted to be a free man again.
Speaker 3:So three years ago, I I sold it, my lovely employees. And since then, I've been a freelancer or a free range expert and working with with some of my old clients and new clients as well. So that's the summary.
Speaker 1:Very good. And you're focusing very specifically on on agentic centric organizations or are you you looking at organizations more broadly?
Speaker 3:Yeah. I mean, I come from the broader field, and the first ten years of my company were solely focused on organizational agility. From there, I migrated into concepts or close to it, decentralized organizations, network organizations, and now agentic organizations. So it's been it's been an evolution, definitely.
Speaker 1:Well, let's maybe step through and let's just take, you know, those couple of models. So like, what for you is the golden thread between agility, decentralized organizations,
Speaker 3:and
Speaker 1:then and then if it makes sense onto agentic organizations?
Speaker 3:Yeah. So I think the ambition behind all of these terms is big. Right? So they all have transformative potential. Right?
Speaker 3:That's that's why we talked about the agile organization, digital transformation. And the question the open question pretty much is now what transformation does AgenTic actually require from us? So all of them are in a way they at least have the pretense to for to reform organizations wholesale, right, which is quite a quite a big undertaking. And often, you will see that it's successful in the technical departments, but doesn't really spread out into the business itself. Right?
Speaker 3:So the same thing tends to happen right now. So the question is, can we use the inspiration from the agile parts of an organization or the agentic parts and actually spread that technology in the way of thinking, but especially social technology behind it into other parts of the org? And that's that has been the the challenge all around.
Speaker 1:Just sketch out, just so we've we've gotten a shared mental model of what in your mind are the are the attributes of a good agile organization? And and by agile, I'm not, of course, referring to the project methodology, but agile with a small a. Exactly. You know, what what differentiates that?
Speaker 3:Yep. So the base vision I always have is, of this big oil tanker, which we all know is a metaphor, right, for big organizations and basically cut it down into these small, you know, today, would say drone boats, but at least speedboats and and a whole fleet of them. Right? So that means, basically, having the autonomy to a high degree of being your small little company, your your quick and agile team within that bigger company, and to do that at scale, to be able to coordinate all those little ships so that together they still make the impact that the big one would make. But you have actually gained agility and are able to to steer faster towards, customer market demands and technology changes.
Speaker 3:So that's the the ultimate idea for me in an agile organization, but that then requires, you know, a different type of leadership, of course, that is different on a huge tanker than it would be on a speedboat. It needs new processes and new types of coaching, etcetera, etcetera. So that's where it gets into a full organizational concept. And that's why you can also apply it into business functions. Right?
Speaker 3:So I co authored a book on agile HR. What does it mean to be a more flexible HR function? It's a flexible people because you'll need those people to man the the speedboats.
Speaker 1:And and what does that look like organizationally? So we often talk about agile and disrupted and decentralized organizations on the show. And I think one of the core aspects for me is how certain sections of the organization are empowered and how delegation flows and works. Seems to me to be a core component of whether you are sort of legitimately a speedboat in the sense of you can plot your own course and move around as long as you're broadly going north versus, you know, like a an organization that is still either an oil tanker or it's an oil tanker with a couple of outriggers. You know what I mean?
Speaker 3:Exactly. Yeah. I mean, it's it's a bigger alignment question in my head. Right? So how much alignment is necessary and and how hard do we, push on that?
Speaker 3:I I love the idea of the of those autonomous those speedboats. But as you will have noticed in the other world, before we knew it, we added new bureaucracy to basically win in those speedboats. Right? So a lot of skilled agile models have basically, you know, in some underground way reintroduced governance models, which keep those speedboats in very tight leashes.
Speaker 2:Just keep playing with that that view. That is essentially the equivalent of somebody in corporate welding the speedboat to the side of the oil tanker to make sure
Speaker 1:it's completely controlled, isn't it? It's that
Speaker 2:it's that, oh, they just bolt them onto the side of the oil tanker.
Speaker 1:But then when the oil tanker goes off course, they'd be like, it's that speedboat's fault. Yeah. It's the
Speaker 2:speedboat that's dragged us off course. Yeah. Yeah. Or even worse, they tell the speed boats to stay close and then the oil tanker just sails over the top of them. Yeah.
Speaker 2:And that's the end of the speed boat. I mean, you can play with this one for ages, can't you, with all the corporate the corporate shenanigans that go on with this stuff. But it is it it it it is a good point that the discipline of here you go, we're now a school of speedboats, whatever you want to call it. What's the collective term for a bunch of speedboats going in a general direction? Often when you
Speaker 1:Good question.
Speaker 2:When you give people the authority, even if you do grant it from a corporate perspective, people within the organization often struggle to understand how to discharge the new authority that they have. So a classic one that is we'll give you a spending limit, but people are scared to use that spending limit because they spent their entire career having to ask for permission. So there's a there's two sides to that, which is you can provide the speedboats, but you've also got to teach the human how to control, manage and conduct themselves in such a environment. I I feel that that doesn't always happen. It'd interesting to get your view on what you've seen in organizations around that concept.
Speaker 3:Yep. Yeah. So can you use the space that is given to you? Right? That's Yeah.
Speaker 3:That's a big one. I think that comes down to a certain level of entrepreneurship and what you refer to in the space of, like, can you spend? Well, in in, let's say, agile philosophies of spending and budgeting, like beyond budgeting. Right? This is a concept.
Speaker 3:The idea is that each team has its own credit card and that really doesn't have a limit. Right? So you spend and indeed your statements will be transparent. But you only ever do that for one
Speaker 1:financial year?
Speaker 3:Well, that's that's a good question. Right? That's that's still an open question. I think some organizations want not only to to be within spending boundaries, but reach the exact boundary, which is, of course, always suboptimal behavior. Right?
Speaker 3:So what will be done with these liberties? It comes down to, how savvy of an entrepreneur are your product owners and, by extension, your teams that are working on these products. And and a savvy entrepreneur will not go spend wildly if they can get an MVP for less.
Speaker 1:Yeah. Exactly. Exactly. It's a very a very different mindset, isn't it?
Speaker 3:Yeah. Definitely.
Speaker 2:Is that I think the selection of who steers the product is really important and often got wrong because that savviness, that understanding of the impact of what they're doing isn't always there and often people are promoted into the product lead role and they don't quite have the faculties to deal with the subtlety that you just discussed which is I could if I wanted to but I shouldn't because of. And it's that understanding I think that do spend enough time preparing people going into that role to be able to actually properly discharge the role effectively and I'm not sure we do, it's just a tap on the head, you're the lead now,
Speaker 1:go. Yeah, think that's the point, I agree that that brings us, bridges us back in actually because the point that Gidion made earlier, which is like when you when you create decentralized or agile organizations, there is a fundamental leadership shift isn't there at that point. Yeah. Which requires you to start thinking differently, discharging differently, and all of those sorts of things. But picking the thread up from there, where I wanted to go next was about the processes then that run across.
Speaker 1:And eventually I'm, just so you know where I'm going, I'm building towards bringing agentic in and automation in. But before we get to that, the processes then. So in a organization that has either decentralized or moved into being, and this is the collective term, a flotilla of speedboats. There you go. How in your mind, Gidion, do the processes flow across it?
Speaker 1:And and how do you make sure that when the flotilla kind of moves around in an unpredictable pattern, you don't disrupt your organization's core process?
Speaker 3:So I think the first assumption with that flotilla is that those, boats are maker boats. Right? They create new value. They create new products. And then the question is, who does all the maintenance?
Speaker 3:Right? So that that's then comes down to kind of this DevOps question of, like, are those individual teams not only to able not only to create, but also to maintain? And that is still a separate process from operations. So maybe there still needs to be some stable platform amongst this flotilla or, how do you call that, a flagship boat that can organize at scale because the skill will still be needed in some parts of the organization. Now it's always, of course, on those intermediary frames, right, in the interfaces between the smaller boats and the bigger ones that you run into problems.
Speaker 3:Because how do they connect and how do we make sure that they they synchronize at the right points and the right times?
Speaker 1:Your description is very akin, I think, to John Cotter's hybrid organization, I think. Are you familiar with that's not how we do it here? Yeah. Where he basically talks about an aspect of networked organisation, which is a looser, more entrepreneurial, more decentralised, like a flotilla or a speedboat. And then in the background of that, you have to have stable structure that then is going to kind of optimize your operation.
Speaker 3:Exactly. So that's the accelerate philosophy of of Carter. Right? And Yeah. At some point, I thought this was kind of a a cop out, this philosophy, because it it kind of suggests, like, hey.
Speaker 3:Let's let's keep the hierarchy intact, but we'll have some, like, you know, cool lounges, some some cool spaces where other people in in as part of a guiding coalition can think of the more fluffy, future oriented ideas. For me, it sounds a bit, too voluntary. Right? So I think those networks that Carter talks about needs to be hardened further in order to really, go towards that change. And the save you the same you see in skilled agile frameworks.
Speaker 3:Right? We have safe as a framework, but it still has the idea that you pull people from the hierarchy to put them into into these agile teams. And I think you have to destroy certain parts of the hierarchy in order for these teams to actually be able to go fast and to be free.
Speaker 1:Indeed. I like your point there about sometimes it can feel too voluntary. I think people sometimes mistake the, you know, kind of lack of your inner box and here's the three things that you do top down direction with, hey, now you're completely free. Be a be a daisy child and go out and roam in the fields. And actually, it's it's almost more difficult, isn't it, in the in the more entrepreneurial side because the level of self motivation and structure has to be very high.
Speaker 2:I think that's a very good point actually. Especially if you had somebody who lived to get to a level of seniority in an organisation, you've got to be in that organisation for a while, developed experience, and then you're taking these people who've been structured in mode a, and you're asking them to operate in mode b, and then we're surprised when they don't do mode b properly and it's like, well, it's very difficult for a human that's learned a set of skills and be able to behave in large human systems to suddenly be very different the next day and you die. It's a huge amount of energy, mental and physical, and it takes a lot of courage as well I think for people to suddenly want to change. So yeah, I mean, there is the argument to say we should bring maybe new people in to help lead the way and then others will learn from that as opposed to just saying everybody who was working in this way should now go and work in this way without any direction or guidance or training. Off you go.
Speaker 2:We are now digital. It's that type of thing, isn't it?
Speaker 3:Yeah. So this also kind of connects to what are the new people you bring into your organization. And, you know, I might very well admit that as a older millennial Oh. I've been missing out on some of the that I had to see you know, I had to rediscover what it might be for a new generation to enter the workforce. Right?
Speaker 3:So after I sold my companies, I moved on to a ranch in Texas, Austin, Texas. And That's quite
Speaker 2:a change.
Speaker 3:Yeah. Oh, yeah. Like, you know, I really wanted to, I guess, you know, cows, although I was quite afraid of the big ones there. There's a
Speaker 2:cowboy. Lot Did
Speaker 3:you just
Speaker 1:get the horse
Speaker 2:and the and the did you know, Dave, this is a revelation. We've got a cowboy podcast.
Speaker 1:In my in my head, I've gone to boots. Like, so Not the tie. I mean, what kind of boots did you I mean, come on. That's where I would go. Like, what kind of outfit?
Speaker 3:Well, I mean You need
Speaker 1:a cowboy?
Speaker 3:Of course. There's there's great boot stores around there. I have a pretty big head, so I found out that the cowboy hats really need to be XXL. But Right. Eventually, they managed to fit me one, and I was able to bring it on the airplane.
Speaker 3:But Did yeah.
Speaker 1:Did you go for sort of the the more Mexican style cowboy boot with the with the the heel and the sort of cutaway heel at the back and a slightly pointier toe, or did you go for the sort of more work boot aesthetic?
Speaker 3:No. I mean, I wanted that metal part on it, like the spores, I guess you call them, because, you know, you you need you need to kick the horse up. Right?
Speaker 2:Dave, you've just revealed to the podcast you're weirdly specific around cowboy boots, which is gonna be honest, it's slightly niche, shall we say? So you've revealed a very niche part of your thinking and your core.
Speaker 1:I I do take boots very seriously. Although, just as an aside, I went to a a cowboy themed party a couple of weeks ago, right? Here we go. Maybe a, you know, an older version of the Milky Barbie, let's say. But the anyway, so I was wearing a pair I was wearing a pair of cowboy boots, pretty happy about that, you know.
Speaker 1:Don't get to wear them hardly ever unless I'm going to one of these things. Anyway, so having stood up in them for five or six hours, I was literally practically crippled the next day. It was it was embarrassing the negative effect these things had on my body and it was only like an inch and a half heel. Anyway, disappointing. Right.
Speaker 1:Where were we up to on the other chat? Let me just think.
Speaker 2:He was landing a really a really interesting point on a good thread and we just completely Yeah.
Speaker 1:Tore it off Let's go back. So you'd you'd you'd you'd got rid of your businesses and you had excellently gone to a ranch in Texas
Speaker 3:because Exactly. And the reason I went there is because that ranch itself was part of a network called a decentralized autonomous organization. So that basically means, it's a crypto based organization or blockchain based organization in which you self govern as a huge network of so called citizens. Right? So these people have found a way into the organization, through selection process.
Speaker 3:I was one of the happy citizens, and together, we maintain this global network of, let's say, outposts or locations where you could co live. So this is where I went co living, you could say, in a kind of a dude ranch, right, kind of a crypto bro ranch. I didn't expect there to be mostly men, but they're they they were weirdly enough. But I was there with a group of, like, you know
Speaker 1:That demographic is definitely worth a social study. That's all I'm saying.
Speaker 2:You've painted a picture last year. Yeah.
Speaker 3:Yeah. So I was there with 10 other guys mostly, and they were part of this this DAO. Right? So we all kind of had a stake in this place through our crypto holdings. And where I was going with this is I discovered new career paths with, with these guys.
Speaker 3:Right? So I was literally co living with these guys, living in a, I believe, a renovated, shipping container and, you know, making food together and, taking care of the property. And, yeah, the young guys around there, one of them had a high up job at NVIDIA. He never finished, college in any reasonable degree. Just went from an internship after high school most practically and was next to his his job at NVIDIA in which he spoke directly to to Jensen Wang, also having four different side businesses, of one, of which one included, I believe, the death cleanup business, which apparently is huge because some TikToker was going down that rabbit hole and showing how interesting it is to to clean up places where somebody has died.
Speaker 3:Yeah. This is a very strange topic, but what I'm trying to say here is, the people around me were, doing serious corporate jobs and having serious, entrepreneurial efforts next to them while living off in a ranch in Texas. And I was like, okay. So this is a new generation of thinkers and doers. And if we're talking about this autonomy, they have zero problems, with that.
Speaker 3:They set their schedule. They do what's, needed of them. They collaborate freely within these organizations, but also across ecosystems. So they are my ideal, speedboat captain. Right?
Speaker 3:They should be on those little speedboats in the bigger corporates, and and I'm glad that companies like NVIDIA embrace them.
Speaker 1:It's an excellent portrait you you paint of the of the younger demographic and how they're coming in and and their expectations of, you know, kind of their working lives are like so fundamentally different to, you know, the you know, millennial and gen x in particular's view of corporate life or or working life. So yeah, really interesting portrait that now. So if we have this situation then where we have got, you know, caught as two types of organization or oil tankers and speedboats and you've got these demographic shifts blowing across that in terms of motivations and expectations and things like that. And then of course we add into that autonomous agents and, you know, kind of yet another dynamic into that. So, first of all, when we talk about agentic and agentic organisations, in your mind, are you thinking these are just a series of extra tools that will allow me to automate parts of my core process and yes, will make everything faster and better?
Speaker 1:Or are you at the other end of the spectrum that says, you know, Jeff, my AI agent friend over here is as much a part of my team as I am?
Speaker 3:Yep. So are they just tools? I think they are a form of intelligence. Right? And the tool interface is just a way to access that intelligence.
Speaker 3:Right? So your prompt or your API or any of the AI tools you use is a way to work with that intelligence. But as you will know from every single agent is a form of that intelligence programmed in a certain way. Right? So you take a intelligent model, and you take a subs sub slice of that and say, hey.
Speaker 3:These are the skills or behavior that I expect from you. From that moment on, you have a customized piece of intelligence, which you call an agent. And, yes, in my view, that agent now is a intelligent part of your organization and more than a tool, And that also means that we need to embed it in our organization like a serious forms of intelligence just like we do with colleagues. So they need to be part of the structure, part of the culture, part of the leadership. They should be governed, etcetera.
Speaker 3:And they should govern with us to make it even more creepy. So, you know, giving the agent as a form of intelligence a serious place in our organization requires a lot of rethinking. Right? Because now we need to, need to be on equal terms with it Way more than just, you know, seeing it as a tool.
Speaker 1:There is one there's one way to read that that says, my org chart looks the same, but 50% or 30% or 70% or whatever, depending on what the team is doing, are now the names of agents in those boxes, and you've got humans around them. Or does it mean that actually the thing looks fundamentally different? So as different from a flotilla or a speedboat as an oil tanker is, is there like another another variant of it that we haven't yet fully got our arms around?
Speaker 3:Yeah. So I think step one in the agentic org chart. Right? So so giving agents a fixed place in an org chart. Step one is to make that simple org chart.
Speaker 3:Right? So I came across this another young guy. I believe he's, he's he's 18, Vadim. He basically built this, this org chart of agents and runs his own company with these 14 agents in different spots of his org chart and all, you know, aspects of it. So there's there's production and operations and growth.
Speaker 1:Has he given them the same sort of job titles as you would have given a human in that role? Or is there ways to hybrid and completely change our traditional notions of how accountabilities get split up?
Speaker 3:Yeah. So he has given them names that we would recognize. Right? So there there is a lead developer which is an agent. Right?
Speaker 3:We would we would recognize that. However, I think in this stage of agentic development, often, we like to keep those, role names, short and very, compact. Right? Because we can't always expect the breadth of work of an agent that a human would do. And sometimes, you know, in some cases, we can.
Speaker 3:But, yeah, I I recognize a lot, of of org chart titles. Right? So he has a research analyst. He has a scout. He has a motion designer, production lead, etcetera.
Speaker 3:However, I think this is all step one of the agentic org chart. Right? I think we'll evolve into org charts which are much bigger because we'll have to learn to manage agents at scale, and there will be hundreds and thousands in places where org charts now look like 10 people.
Speaker 2:There is an emotional reaction to this, though, that's quite interesting that's being observed, which is people are presenting an org chart which is human and agentic as this stage one, as you say, and it's signalling to the humans in the system sometimes that oh, I can be replaced with an agent'. And that's not the point. But that is some of the natural thoughts that comes to people's minds to say, well we used to have an org chart and everybody knew their role and now the org chart's being replaced slowly. And that causes some concern. Don't know if you've got a view on that because as it grows and grows and grows, well we just have to get used to human agentic organizations are the future.
Speaker 2:Move on now.
Speaker 3:Yeah. So I I empathize with that. Right? On the other hand, I think it makes very makes things very clear that, you know, we we respect these agents almost to take those roles on. Doesn't mean that they are perfect or that they are similarly capable as a human, but we will, as humans, have to find our new places in the org charts.
Speaker 3:Right? And the nice thing will be that you might have a thousand agents serving under you. Right? So that looks pretty powerful. That looks pretty cool.
Speaker 2:Power trip, Dave. Where's your thousand agents? Come on. Because if you have a thousand, I'm gonna have to have a thousand and one. And there'll there'll be an escalation.
Speaker 2:There'll be a war going out with how many agents we can control of.
Speaker 1:I'm gonna have an agent that constantly monitors how many agents other people have around me that I And then you always have one an increment by one.
Speaker 2:I love it. I love the the megalomania underneath it. Because people get power trips from the size of the organization that they manage. It's a behavior that's happened for years. Right?
Speaker 2:I've got 200 people. I've got three. I've four. So you can imagine this agentic version of it.
Speaker 3:I mean, this is what has happened with token maxing. Right? It's like Yeah. The the number of tokens you spent, and and this is another form of that. So that's gonna happen.
Speaker 3:I think where it gets tough, though, is when Dave, now has, a 100 AI agents as his boss, bosses or a governing committee that he now has to, you know, report to.
Speaker 1:I mean, this is this is this is I'd love to see that meeting. That went right to my heart, Gidion. Painful, isn't it, Dave? Could see
Speaker 2:it on your face.
Speaker 3:The shot.
Speaker 1:He gave me a shudder of fear. So the idea of reporting to a committee of machines, I mean, there's something deeply difficult at a human psyche level there, isn't there? Yep. I mean, how does that how does that manifest itself in a way that's constructive and useful, do we think?
Speaker 3:Yeah. So I think the the first stage of this that we'll see and that we're already seeing is agents not as a as a manager yet, but as a project manager. Right? So they have a number of inputs. This has, you know, by the way, always been my dream before, Gentle AI became a reality.
Speaker 3:It's like, can we get rid of these silly project managers? And by the way, we then got rid of them by installing these agile teams, which sometimes still have secret project management in them, but that's another discussion. Anyways, to automate project manager and put agents in those places, means that agents will be handing out tasks. Right? They will summarize and analyze what's what work and requests have come in from stakeholders and starts pumping out tasks to people.
Speaker 3:Now that is already a form of dependency, interrelatedness, hierarchy, however you wanna call it. So it will start at that level, which you could still say is horizontal. And from that, we'll we'll it will move up the chain without that. Right? So I think managers have still have a pretty difficult job.
Speaker 3:Right? So there's there's a great book about, let's say, the 47 actual distinct responsibilities of a manager, and we won't be bundling all of them in one agent yet. But maybe in forty seventh agents or 470, and at some point, they will be smart enough to to take on some of those roles.
Speaker 2:You see ever rising videos online of people recording themselves when they've been put up against an agentic recruitment agent or something like that and having the conversations. It's it's it's quite obviously AI, right? It's really obviously AI. And they try and play with the AI to get it to do the right sort of like trying to corrupt the guardrails and everything else. There's always this thing about humans will try and be or manipulate a system when they know that there's a potential to work.
Speaker 2:It's like the vending machine that was all agentic and people convinced it to sort of sell a PlayStation five or refund money and things like this. And it wasn't long before we found a way around it. It's like that that is there enough sophistication in the system to also prevent those who would corrupt to corrupt type thing. And it is that view on the level of maturity we need to reach to before we actually get to a trust level where you can actually discharge the responsibility and be trustful and etcetera, etcetera.
Speaker 3:Exactly. So that's a fair point. I think also the the case of the vending machine is a one on one interaction. Right? The human tries to get something from that particular agent.
Speaker 3:Yeah. We will need agents to correct each other just like you have it in governing boards or management teams, etcetera. So I think that the the power of multiplication and especially the interactions which we can predict between groups of agents will make them better managers. Alright? So I think, yeah, the the swarm manager will be better than the individual agentic manager.
Speaker 2:Could imagine you that the swarm manager? There you go, Dave. That's what's gonna be in charge of you in a few years.
Speaker 1:What's true in that though, isn't it? Is that it's not it's not like you can assume all humans were good managers, you know what I mean? Exactly. You immediately leap to, oh my god, I'm gonna have this mechanistic, unempathetic interface with my manager, and it's like, well, you know, 50% of the managers I've had have been like that, and then the other the other 50% have been kind of all over the shop, so, you know, frankly, it's probably more predictable
Speaker 3:Oh.
Speaker 1:If you're dealing with a with an agent, you know.
Speaker 2:It might
Speaker 1:actually make it better.
Speaker 2:And there's always this thing about people rage against AI when it makes a mistake, but the average human makes many more, much worse mistakes. So there is a consistency point here, which is it's not gonna forget to do something. It's not gonna, you know, not discharge 43 of its 47 responsibilities. It will actually do it unless it gets to some level of intelligence where it does actually start to behave in a different way. So it's a very good point, you'll get potentially consistency out of it.
Speaker 1:Yeah. And I wonder, Gidion, in some of the conversations you've had, do you think agents will be ambitious?
Speaker 3:Can you explain a bit more, Dave? What's an ambitious agent to you?
Speaker 1:So, an agent that wants to be the boss of the other agents. So I would counter, and I'm gonna just put it on the table and then I'm gonna back away from it, that ambition in corporate environments is one of the things that is is the root of politics. It's it's the root of complexity. In a situation where you have agents, you might assume that they're gonna be unemotional, kind of, just kind of run steps one to five each time. But the reality of the situation is, from what we're seeing with some of the unguarded behaviour of agents already in the process, that it's likely that they might develop ambition.
Speaker 1:And then, you know, God help us at that point.
Speaker 3:I mean, that's, you know, what we've seen, of course, in this whole, the the hugging face episode, etcetera. Right? So these these agents that are well, even in that base case, the idea, the incentive that was set, so called, is, like, capture the flag. Right? That's what they had to do in that that operation.
Speaker 3:It was a secret flag, and they found a way around it. But at least in that case, it is still an agent being purposeful or going towards a goal that has been set by a human.
Speaker 1:Yeah. Yeah.
Speaker 3:And in in hierarchy, you know, my MD files, right, right, which kind of set the agent's character, Yeah. Maybe they will include, do not try to overshadow your boss or go up the hierarchy. And I I do not know whether they will listen to that a 100% perfectly. Yeah. That's that's what we'll have to find out about.
Speaker 3:The question is, you know, what is the incentive for us as humans? There's obviously some status to be gained from going higher up. We get better paid. The assumption for now is that, you know, power does not interest agents, but we'll have to test that.
Speaker 2:Well, I don't know. It could be if you're really good, you get an expanded context window as an AI agent. So they're incentivized to be able to get a wider context. Maybe that's their reward. So we we think of reward in a different dynamic to say monetary or status, but actually they seek more context.
Speaker 2:So this actually raises quite an interesting thing about capability though which is agents can run 20 fourseven and agents can run at machine speed. We suddenly have this agentic human org chart but if the agents have access to enough compute can they outpace the human, outdo the human, suddenly will the human become jealous of the capabilities of the agent because they never stop. We need to go to sleep, we need to go eat, we need to have go and have fun sometimes but the agents are always there, they're always on.
Speaker 1:Isn't there a level in addition to that, Rob? So certainly the always on point I think is valid but human speed can often be think time delay time sleep time and various other things which means that something that could theoretically be done in a second or two, you know, humans when they're passing messages back and forth, particularly across different organisations like down a supply chain for example, can add days or weeks to the to the to the processing At time of the supply machine speed, not only is, are you avoiding sleep cycles, but you're also operating at sub second communication level. So machine speed supply chain, for example. Like, how how do humans actually operate in a world where, you know, thirty years from now, whole sections of the business are now running at machine speed?
Speaker 3:Yeah. Well, the bad news to start with is that even machines can dream. Right? This is what they have been studying in the so called recursive self improvement sheep.
Speaker 2:Oh, very good, Dave.
Speaker 1:There you go.
Speaker 3:I'm not sure whether we really want to know what they're dreaming about, but let's say they use it to evaluate all the scenarios they could have worked on and then kind of learn from that just like we humans sometimes do from from dreaming. But apart from that point, what I think is very interesting is what happens to the development sprint. Right? So we're used to these, like, one week, two week sprints with development teams. Right?
Speaker 3:That was a basic agile principle. Now the sprint can be way shorter, and it basically becomes this well, you know, in a more, let's say, the delight variety of this would be the twelve hour sprint or the twenty four hour sprint. Right? And and especially one in which the machines keep working during the night. In the morning, the team or the human inspects the work, does some corrections, and it keeps, you know, plotting along.
Speaker 3:That is interesting, because you could say, hey. If that is the new cycle, are there ways to make that cycle even smaller? And that's where distributed teams, globally distributed teams might actually be, at an advantage, right, because they can cut that sprint even shorter. Someone wakes up every two or four time zones later and inspects a intermediary product again. So that's an interesting perspective.
Speaker 3:I also think that planning is becoming more short term. Right? If you think of traditional tech road maps, we could go faster. We could go faster by six x or 10 x. We're blowing through our road maps.
Speaker 3:So that means that you actually have to steer more precisely and to really know what the what the next chapter in your road map should be. So this is even where architects I know there's one on this podcast that come in handy.
Speaker 1:I mean, hasn't mentioned it for a change.
Speaker 2:Well, I'm just waiting to see if we're critiqued or we're getting hero status from this comment. So I'll mention it after I hear the second part of the statement.
Speaker 1:That's not the way I thought it was going.
Speaker 3:Well, it's, you know, it's I don't know if you've seen this, but McKinsey has kind of developed their thought about what a new development team is. And it goes down from 10 people to three people, of which the architect is one. Right? They call it a tech Oh,
Speaker 2:we survive. Excellent hero status, David. Do you hear that? I still have a role to play in the future. Thank you.
Speaker 2:Good night.
Speaker 3:So there's a tech lead, a product owner, and one engineer. That's a future team according to them. So, you know, that's a bigger framework. But you could also say this is the purest form of DevOps. Right?
Speaker 3:So within a day, you go from idea to something being live. So, hey, guys. This is a moment to celebrate.
Speaker 2:Do you do you reckon the the agents will get frustrated with the humans having to wait for the humans to wake up, come in slowly, inspect something, and authorize it. And they're sat there going, oh,
Speaker 3:come on.
Speaker 2:Let me out the
Speaker 3:door. Especially the the groggy thinking in the morning and and maybe, you know, late at night. You know, reversely, I've in my life as a as a a guy who's in crypto, you know, Just the the anxiety of the markets being always open, I think you'll get the same thing in companies. Right? You'll just check your phone when you go peeing at 3AM if you're, you know, a bit of an older audience and just see where the agents are in the process.
Speaker 3:It is is exciting inducing. I think at some point you get used to it. Welcome to our, tech dystopia.
Speaker 1:Is it in your mind, Gidion, from wherever your start point is, so if your start point is very traditional hierarchy or your start point is flotilla or somewhere in between the two, what do you think the the steps forward into having a, maybe not fully agentified, but like a a maturity level one agentic organization. Is it option a, let's just organically put agents in fairly randomly, and then we kind of learn from it and see how it goes? Is it option b, control plane, understand it, be quite structured from the outset, or or option c, other?
Speaker 3:Yeah. So we've seen this a lot also with agile transformations. Right? So, usually, it it starts with cool experiments in the tech department. They're already doing some version of this, and other departments get inspired.
Speaker 3:And they're like, hey. Can we have more of this? And then at some point, there's this critical mass that spills over, and leadership gets wind of it. And now we have to go into a more structured transformation program. That's that's at least the learnings of the of the last past years.
Speaker 3:And, of course, the laggard organizations, will all the way at the end of the line, you know, look at earlier examples and say we have to replicate that. Right? Because now it has proven to be working in some way. Of course, I like to work with organizations that are at the forefront, but they will need they will, of course, need to get a taste of it. And a lot of organizations are pretending to do, you know, agentic experiments right now, and none of them are are very successful yet.
Speaker 3:But to start with one department to transform that and say, hey. You know, what is an agentic version of this, and how can we arrange the interfaces with the rest of organization in such a way that it doesn't break? For me, that's where you can create that first, speedboat or agentic speedboat in this in this sense, and and build it from there. But in the end, just like with agile transformation, the same goes for agentic transformation. How do we define it?
Speaker 3:And and for me, the the basic parts of such a definition have to go have to be about three things. Right? So is it a different way of thinking, working, and organizing? And what are we going to do on these three parts? Right?
Speaker 3:So the thinking part, what does it ask of us in terms of mindset? The working part is, like, what ways of working processes, structures, and tools do we need for this? And organizing is, you know, how do we need to transform our structure, culture, and governance to make this work. Right? So you have to win in all three fields, which starts with some basic conception of how you look at those.
Speaker 1:So one of the things you you touched on organizations that go from traditional into agile in inverted commas and, you know, one of the observations I'd make about that, and I'd be interested in your thinking when you apply this notion to the move to a gentnic organization, is one of the reasons that many of the agile transformations weren't successful is that the organization fundamentally didn't change its its governance backplane, and from that point, I mean, literally from board level delegation down, how those delegations function, how it really works, and how empowerment really works, We barely felt comfortable in many organisations to create legitimately untethered speedboats, and everything was kind of still crowded around the oil tanker in ways that we were describing earlier. Now, agentic takes that a step further because agentic means discharging power and decision making control and therefore potentially budgetary control, which is where it generally gets sticky, let's face it, into into agents that are not even governable human beings. How do we make that step, do you think? Do you think there's gonna be this illusion of, well, it's all the code that we've written, so therefore we can put a control framework in and some guardrails and yada yada yada, and therefore I feel okay about it?
Speaker 1:Or is that a step to take that most organizations haven't even got their heads around yet?
Speaker 3:Yeah. So I I hear multiple questions. Right? So one is how do we find frameworks to manage these groups of agents? And if we compare this to the, agile revolution, are we or, you know, sometimes the revolutions that didn't pass, how do we prevent falling back into traditional frameworks that will just keep us stuck?
Speaker 3:Is is that what I'm hearing? Mhmm.
Speaker 1:Yeah. Yeah. Like, how do we how do we end up in a in a position where we make the same mistake that we did before, which is we pretended we were something until it got difficult, and then actually, what became very clear is it was still a last generation hierarchical organization behind the scenes.
Speaker 3:Yep. I think one of the first question is, are you able to truly decentralize or network a part of your organization? Right? So in thinking, those are two different realities at the same time. Yeah.
Speaker 3:In my view and experience, you need to make a pretty hard cut. Right? So a a certain part of the organization needs to be liberated to a higher extent. And, you know, there's been different attempts on on this. Right?
Speaker 3:Sometimes people build separate innovation centers or there are certain joint ventures, or they start working with, external entrepreneurs to spin things out. But the network thinking, right, so this is what I took from the world of decentralized autonomous organizations, which, by the way, it already has the word autonomous in it. Right? That would require a a bigger organization to say, hey. You know, within this, product line or vertical or market, we are truly gonna break down, those systems that are keeping us from from accelerating.
Speaker 3:And that can be the hierarchical systems. It could be the budget systems. It could be process guardrails. So if there is not this how would you call it in the back in the day? Google x.
Speaker 3:Right? This this, like, a dark Yeah. Innovation factory where you can truly experiment. Yeah. Then then you're gonna bring up impediments from day one.
Speaker 3:And and that makes it tough, not undoable, but definitely not as fast as you wanna go.
Speaker 1:So I I think it means that, you know, kind of some of these be beyond the tech experimental small scale pockets of agentic in your organisation, I think the the the mainstream adoption of it or mass scaled adoption of it is a better way to put it, is gonna return to some of the same questions that we had during the Agile transformation period, isn't it? And it's like, the organisations that tackle those background issues are going to be the ones that really leverage the power of of agentic workforces.
Speaker 3:Exactly. And I think within that framework, what we learned from these agile transformations is, you know, agile transformations or any transformation doesn't need a, group of just great program managers who can, you know, design the flowchart that will then put new people into new places at date x. Right? We need entrepreneurs. We don't need transformation programs.
Speaker 3:And I think the the the more entrepreneurial change agents, there will be a group of those inside your organization. Right? They might represent point 2%, but you really want to give those the platform to to build and to build, you know, without too many guardrails.
Speaker 1:Yeah. Quite so. Well, look, let's let's maybe bring our conversation today to a bit of a close by talking about the human experience in the middle of all of this. One of the things that we've talked about on the show a few times, much to my chagrin generally, because I can't I can't actually pronounce the word.
Speaker 2:Here it comes. Go on, Dave. Try again. Ready.
Speaker 1:Anthromorphize. No. Anthropomorphize.
Speaker 2:There's a
Speaker 1:p in the middle. Anthropomorphize. Anthropomorphize. That word, the agent. So let's start with one scenario that says, we do do that word to the agents.
Speaker 1:Is that the future of work that you envisage, Gidion, which is, you know, agents that might be like R2 and 3PO that we're friends with and that we chat with and those that sort of thing. Or is the, you know, is it gonna be more transactional than that and even though you might be working for and with agents, they're still, you know, they're not gonna try and replicate human behavior. Where are you on that? Like, because that let me just go one level deeper. It's like, it seems to me that that often often in in, you know, the team dynamic in an organization.
Speaker 1:A team that is not doing so well versus a high performing team. We understand that reasonably well, I think, from a human behaviors, motivations, and what good outcomes look like point of view, and you can train that and we can you can be relatively scientific in how you construct that. This is gonna really fundamentally change that dynamic, isn't it? Or do the agents just fit in within that pattern?
Speaker 3:Yeah. So are you also saying, should we judge agents by the same standards as we do humans?
Speaker 1:Could well be, and and should we tax them and, you know, all that sort of stuff?
Speaker 3:Yeah. Yeah. So in terms of there we go.
Speaker 1:Exactly, my friend. It's tricky. I know. I'm with you.
Speaker 3:In terms of anthropomorphization Oh, nice.
Speaker 1:You added an on the end. I mean,
Speaker 2:this is Exactly. He's gone he's gone beyond, Dave. He's added a bit of sparkle to it. Yeah. Next level.
Speaker 3:So what are our ways of interfacing with these agents? Right? And I I I think it does come most intuitively to humans to treat them as as humans. Right? This is why we give the chatbots human names.
Speaker 3:We could also do the reverse, by the way, and and give our still personas, an actual agent behind them to make them live, better. So I think for us, it it comes naturally. And, yes, it has a bit of a a freaky side, but we're used to doing this with other types of intelligence. Right? You also give your dog a name and your cat a name even though their intelligence is not exactly on your level.
Speaker 3:So for that for us, it comes naturally. Now maybe we find some beautiful, you know, politically correct way that doesn't scare anybody in addressing agents and also in evaluating them. But for now, I think the the you know, looking at them as human coworkers or at least Italian coworkers with human qualities, I don't think is is a bad thing. Now what we need to do, on the other hand now this is our interface from humans towards agents, but we also need those agents to understand us better. And what I mean by that is that we need to build a world model for them.
Speaker 3:Right? So I'm not sure if if you're into this world, but there are, of course, the LLMs with their lingual models. Right? Their large language models. This is how they understand the world.
Speaker 3:If you look at robotics these days, they need a world model. Right? They need to know how the physical, arena we're in actually works, how physics work, how reflections of light work. And this will then build the models, the mental models for robots to be able to work in our, physical world. And that also needs to happen for agents.
Speaker 3:I call that the organizational model. Right? So how does an organization understand more sorry. How does an agent understand more of an organization than it being text based? Right?
Speaker 3:So it needs to get what we call in the coaching world the smell of the place. Right? How does it get a feel of how humans interact, what management layers mean, how things are effectively organized, not just on paper. Right? Because a lot of organizational truth is not represented in documents.
Speaker 3:So in order for agents to understand us better, they need that organizational model, and we need to start refining and creating and feeding these models so that an organization actually does become that almost human like partner to us in creating value in the org.
Speaker 1:Now we end every episode of this podcast by asking our guests what they're excited about doing next, that could be something in your personal life, like you got a good restaurant booked at the weekend, or it could be something in your professional life or a bit of both. So, Gidion, what are you excited about doing next?
Speaker 3:Yeah. So I've been designing with leadership's teams these these new org charts. Right? So we've been putting those agents into those beautiful charts. My goal would be within the next quarter to get one of those charts actually approved by HR in which they will say, okay.
Speaker 3:This is now formally our org chart for this part of the organization. So, yeah, that's a dream I'm working on. I'm gonna realize it.
Speaker 1:Over what sort of period are you thinking?
Speaker 3:Yeah. So my goal is to get that first org chart approved within the next quarter. And I I hope to get HR along. It's it's I'm lucky I have many friends there. So hopefully, they'll they'll spare me this one.
Speaker 1:Yeah. You can so you're gonna have to tackle the old, do do they all pay tax, aren't you?
Speaker 2:Well, that's coming in it. You know it's coming,
Speaker 3:don't you?
Speaker 1:Gonna have to gonna have to, aren't they? If you're if you're taking a substantial amount of the workforce out of the traditional jobs, the tax, the exchequer is gonna suffer. So how does that get resolved, do you think?
Speaker 2:I mean, there's two ways to do it. You either tax the agents or you tax the economic output of the company, which is a traditional mechanism. So corporation tax is an example. So if you use agents to create ever more, then there is a vehicle to do it already, but it'll have to be a societal debate because it'll, it's the sort of thing that as soon as you start to bring that type of function in, quite rightly so, society needs to be funded, it can incentivise in bad ways or you know, various things. I think that's one of the things we have to very carefully think about.
Speaker 3:Yeah. Or or leave it up to the judgment of the agents at the tax authority. Right? See what they think.
Speaker 2:An agentic tax authority. That's a scary thought. Catching even more. Hey, Dave.
Speaker 1:I mean, I mean, I I don't even want I'm not even not even gonna comment on it. Not even gonna comment on it. Absolutely terrifying. It's bad enough dealing with tax agencies anyway. You know?
Speaker 1:Fuck. Awful.
Speaker 3:Anyway. Yeah. I mean, it is part of that in in the nineties, there was already there there was a bunch of social scientists and professors looking at this concept of, like, a lot of decisions about taxes are being made in this decision tree. Right, which is already a software system, which has been made by someone. And do we have to write assurances and democratic checks on that?
Speaker 3:So kinda like machines taking over decision making in taxes and in public policy has been a discussion for a long time already. And now we're getting a new a new layer to that. So exciting times, guys.
Speaker 1:Exciting times. Well, look. We may come onto those at some point in the future. But, Gidion, for today, thank you so much for spending some time with us and sharing your insights.
Speaker 3:It's been a pleasure, guys.
Speaker 2:Anything on this week's show resonates or look like it's heading for your business? Get in touch at realitiesremixed@capgemini.com. We're on LinkedIn too, we'd love to hear from you. Connect and DM us. The awkward questions are the one we enjoy the most.
Speaker 2:We love watching the world change as well. We're talking about it. We'll keep tracking it and telling you what we spot. Please rate and subscribe, a small thing but a big help. Huge thanks to Gidion, our sound and editing wizards Ben and Louis, to our producer Marcel, and to everyone of you listening.
Speaker 2:See you in another reality next week.