E172 - Scaling Smart In The Age Of AI - Helena Hörnebrant

We explore how to scale in the age of AI without drowning in complexity, moving from one-off pilots to reusable platforms, and from efficiency-only thinking to adaptability, trust, and aligned autonomy. Helena Hörnerbrandt shares enterprise-tested principles that travel across 65+ markets, plus a real ABS rollout that shows product over project in action.

• deepfake demo as a visceral risk signal
• value-led scaling and decoupling growth from OPEX
• escaping point-solution spaghetti with use case families
• bimodal model for value creation and capitalization
• governance that speeds decisions, not stalls them
• continuous design over static process change
• trust, discipline, aligned autonomy, and mastery
• product vs project and market enablement roles
• ERP standardization vs data pattern standardization
• the intelligence plane and shared, synthetic, real-time data
• buy to build, ecosystem scaling, and partnerships
• Europe’s need for sandboxes and co-created guidance
• AGI outlook across both risk and opportunity

Follow us on youtube: https://www.youtube.com/@aiawpodcast

Chapters

• 0:00 The Deepfake Wake-Up Call
• 3:43 Why Experience-Based Learning Matters
• 8:55 Defining Scaling Without Growing Complexity
• 15:55 Escaping Point-Solution Spaghetti
• 20:34 Use Case Families And Portfolio Thinking
• 26:50 Governance For Speed, Quality, And Value
• 34:20 From Efficiency To Adaptability
• 42:00 Ecosystems, Buy-To-Build, And Partnerships
• 49:15 Enterprise vs Local: Scaling Across 65+ Markets
• 57:30 Carve-Out To Multi-Brand Platform
• 1:06:10 Principles: Systematization, Design, Trust, Discipline
• 1:17:20 Aligned Autonomy, Feedback Loops, Mastery
• 1:28:30 Product vs Project: Scaling ABS As A Case
• 1:37:40 Platform Patterns vs ERP Rollouts
• 1:45:30 The Intelligence Plane: From Internal To Shared Data
• 1:55:10 Future Outlook: New Objective Functions
• 2:02:40 Europe’s Path: Regulation, Sandboxes, Labs
• 2:12:10 AGI Spectrum, AI For Good, Closing

Transcript

Helena Hörnebrant: 0:00

To see how fast you can do them so people get more respect towards what they read and see and use. And we have had um that uh showed to us so 10 seconds of his voice or 10 or 12 seconds of his voice, and then just another public uh video of him and what you can do with that. And I mean, probably uh if you're super super close to him, you can really hear the attonations, but imagine if you only get five or ten seconds of that, and if you're doing a lot of other things, of course you're not gonna start to react the first thing you do.

Henrik Göthberg: 0:43

And uh what was the deep fake that you exposed him to? Can you can you can you share the uh this because I thought it was pretty funny.

Helena Hörnebrant: 0:51

Yeah, so uh he was basically uh saying hello to everybody, and then uh basically telling the employees that they will get fat bonuses, and this is in terms during a time where we do more savings than ever.

Henrik Göthberg: 1:08

So yeah so but just to prove it uh something that is believable but crazy, and then how fast and simple it is, because it really means then this threat surface area is huge, yeah, because anyone can do it. I mean, like to keep to imagine different ways that this can be exploited when it's that simple, yeah. So and and what was the you know, what was the outcome of the conversation or the uh you know how what was the you know, where did it land?

Helena Hörnebrant: 1:35

Of course, people, even uh my tech people who is really well uh skilled still was very surprised how easy it was. So I think uh when people talk about it, it's one thing it becomes very abstract, but when you see it and at the speed, you understand how dangerous it can be unless you you you pay attention to the critical content.

Henrik Göthberg: 1:59

I also get another sense from when you do stuff like this, why and why experience-based learning is so important in this context is that even people in our staff that are supposed to be data engineers and they're working on on the normal IT stuff and they're really, really skilled in all that, there is still a divide in terms of the the cutting-edge AI and how to work with that. And and um, so even even our skilled engineers that do our day-to-day data stuff can be humble and should be humble because what we are getting into is like, okay, it's not only about this is security now, we need to manage that IT security. What is that all about? We don't really know, we need to learn, yeah. And it also means for me the broader context, what and how can we use AI anywhere? If if it's so easy to do this part, you know, is it is it us that also need to be very humble about this is new learnings that we all need to go on a learning journey?

Helena Hörnebrant: 3:01

I think so. I think the whole AI is a human story and it's about learning, it's about exploring. And I also think that the time where we can be an autopilot is almost gone. Like we have to be in connection and uh beyond when we see messages.

Henrik Göthberg: 3:18

So beyond in terms of what you mean. Like uh we need to all this that is happening around us, we need to actually pay attention to it to quickly, oh, is this something I need to internalize now? We cannot just be an autopilot in relation to what we are the old way of working. Is that what you meant?

Helena Hörnebrant: 3:35

Or yeah, exactly absolutely. And I think whatever we thought was half good a month ago, give it six more weeks, and it's not gonna be half good. It's gonna be good. Maybe not good enough, but good. And then it's gonna be good enough, and then it's gonna be super good. So the the this is a growth journey that we have uh in front of us. I think everybody has that.

Henrik Göthberg: 3:56

So it's a it's a it's a personal growth, it's a team growth, it's it's organizational growth where we need to be very uh open to that what works now or what doesn't work now will work very short in a very short time. And that's and one of them, or many of them will change the game when you start doing it at scale.

Helena Hörnebrant: 4:18

Yeah.

Henrik Göthberg: 4:19

This is a great uh uh segue into what we're gonna talk to about, which we we we put the theme um uh scaling business in the age of AI. And uh, you know, uh this is one of those uh podcasts that I've been thinking about doing with you for years. We know it we go way back. I'm gonna introduce you now, Helena Hernebrandt, and we've had an ongoing conversation around how to drive change at scale since I don't know, 20 years maybe in different ways. I mean, we've been we've been in different roles in different uh parts of our lives, but that question has always been there.

Helena Hörnebrant: 4:57

Always been there. How do you scale with people, tech, and yeah, enterprise or small?

Henrik Göthberg: 5:02

Yeah. And um so with that, I want to formally welcome you, Helena Hörnerbrandt. And I think we need to start with putting your sort of a little bit of background in relation to scaling and all that in the context. So, you know, could you just give us sort of the five-minute version of who you are? You know, what are you working on now? And a little bit like I mean, like you've been working on this from UEM to Nasdaq to uh to to Scania to Troton now. Yeah, so many different angles.

Helena Hörnebrant: 5:29

Many different angles. But first of all, thank you for giving me the opportunity to be here uh to sit down and unpack some things together in a conversational forum. I think that's super important. That's how you learn more most of your daily work and to to really stay ahead of things. Um, so I'm the chief technology, I'm not. I'm the chief information officer and chief digital officer, so CIDO at Hoton Financial Services, which is really uh it's really the financial services of Trayton Group. And Trayton Group is the commercial vehicle site of Volkswagen, and we have a number of brands, Scania M IN, VW International, and next day right in China. And uh my role here is of course to lead digital transformation, data, uh, making sure that we scale the possibility of how we use uh data and data-driven solutions, how we lead growth in a financial ecosystem. Um, and and I think uh why I love this type of job, or the change of it, or the growth of it, or the scale of it, is really uh that we get to use technology to empower people. And that means that we need to strengthen trust between people with technology or data, and uh, we need to also be much more purpose-led than we have been previously, and I think also intentional given the speed of the technology.

Henrik Göthberg: 7:05

But just you okay, so you're CIO CDO now in Troton. I think you had a COO role before, uh before uh in the in in when it was called Scania Financial Services, or is that is that not the case?

Helena Hörnebrant: 7:24

Uh no, uh in uh Scania Financial Services we never had a COO role. Oh, you didn't have the role that I no, we had operations and operational topics and all of that.

Henrik Göthberg: 7:34

Okay, that's my mistake. So so you you had a slightly different role, and then before that, uh I think you've been working in in several different functions with how to scale up things.

Helena Hörnebrant: 7:45

Yeah, so my function at Scania has been more transformation and operations, uh, and but very much uh in the heart of digital. And then uh also had the opportunity before that to be uh heading the PMO at our former CFO or former former CFO, used to be the CEO of uh Trade on Financial Services, and then um uh I really learned how to scale en masse and really figure out how to work in an extremely distributed company. I mean, we are in that sense very unique. So we are a large enterprise and also with a bunch of smaller entities, uh, a few large entities as well.

Henrik Göthberg: 8:32

So this is going to be something we can circle back to during the pod, you know, the the flavor of scaling something enterprise grade and where we has where it has a big center in one way, but at the same time is very entrepreneurial and distributed in terms of a very basic, simple market out there, and we need to make it work for everyone.

Helena Hörnebrant: 8:51

Exactly. And I think uh the reason I could take on such a big challenge was really my groundwork I did at Owam, OMEX, NASDAQ, and then also running a few of my uh entrepreneurial initiatives um uh to learn about that it this is always people, people, people, people, and then a bit of tech, and then you can empower both an organization and a company with that.

Henrik Göthberg: 9:20

So very quickly, and in this is back in the day when you work in OEM, you what was the work then? Because I think that sets you up in many ways.

Helena Hörnebrant: 9:30

It set it set me up in many ways. So I came in just when they formed the name OM. So it used to be Alpha Bay, and then it was OM, and then it was OMEX, and then OMEX Technology, and then Nasdaq Omics. And I worked on the institutional banking side of it. So really starting uh my first job to see how we can get a paper stock into fully automated flow and really learning at that time, already at that time, how important the data was. Yeah. But also that speed matters and security matters, yeah. And that to scale with speed, security, and still within good, healthy margins.

Henrik Göthberg: 10:13

Because I remember this is back in the day, but you because you were literally going around the globe at some point and working with these projects, almost like my understanding from it from the outside, like okay, now we have a new stock exchange, we have uh something of we have a new uh client of us that wants to use our technology, and we need to onboard them into at this point in time, very sophisticated ways of working compared to where they were at.

Helena Hörnebrant: 10:37

Yeah, exactly. So uh we have had a few institutional banks where we have really gone from fully manual to fully automated and what that means. And in the very beginning, it was more like, can I trust the system? You know, it was really funny at that time. Today we laugh at it, but that was the reality at the time. And and I think uh the most interesting thing was that we cannot work uh like the Danish does, or we cannot work as the French, or the English people couldn't work like the New Yorker. And today it's so standardized, so harmonized everything. So now we argue more about how many uh times do I have to um use the actual um um numbers and figures, and really pinning in what I need, how many times do you need to pin down that? Uh, and how fast can you get to the stock exchange to broke a service? So it's really about milliseconds, it's even more than that, and the infrastructure and how close you can be to the service and all of that.

Henrik Göthberg: 11:39

So has that changed a bit?

Helena Hörnebrant: 11:41

It changed a lot. Speed in this world is something completely different to speed in the world that I'm working with at the moment.

Henrik Göthberg: 11:50

All right, so nice introduction. So, with that sort of framing, let's let's sort of look a little bit about putting a context on the theme as such. So I was thinking about when you're talking about uh scaling, or I I even we could um think about the word industrialization as well, but I think we use scaling. So when when you are trying to describe that uh as from your CIO CDO role and discuss this in in the current management team, because then you have you're part of the management team within Trouton Financial Services. If you if you go a little bit like to highlight how that looks like, and then how how would you describe and discuss these topics then if if we used start broadly, like well, if if let's start with what I defined as scaling, because I think that's where it starts.

Helena Hörnebrant: 12:41

Scaling can be done in so many different ways, but I think it's um how do you grow and um really grow impact of the business without growing in complexity? Because it's easy to just add people, systems, companies, that's just growing and growing with complexity. But uh at a certain amount of time or uh size or complexity, you you can't really steer that in a systematic way. So I think scaling in a good way is really about creating leverage from your existing position and see how you can add off without uh doing um too much, adding too much complexity. And of course, in the beginning, you might actually add on complexity, you need to start somewhere, build a foundation, like you build a house. You you need the base to be stable. Then you really need to look at how you are optimizing and scaling as the next uh or expanding, as we would say, to into services.

Henrik Göthberg: 13:49

If I if I let me test, if I there's there's a couple of different ways how you can sort of lean into that definition. I let me I will try one now and see if we're on the same path. So one story that I I think we're coming back to all the time, which I think is about this, is how do we decouple revenue growth from OPEC spending? So basically, we want to build something that can take on more and more revenue. We can we can do more and better deals, but every time we take on a new market or we take on a new, you know, new 1,000 more customers, we don't want to have a linear cost expansion of our OPEX. So this is what I mean. Like scaling for me or industrialization for me, what you are describing now is how can we get to a point where we scale and get more revenue, but the complexity doesn't uh scale, and therefore the costs can be is not uh following, or even to the point where it can go worse. That the more complex you are becoming, the cost is actually going out of hand.

Helena Hörnebrant: 14:52

It's going out of hand, and I think uh this is a balancing game because you if you want to move fast, you can't do everything at the same time. So you you really need to build a fundament, but you need to fastly move over to how you can build it in a more scalable and less complex way. And I think what we need, what you're saying, if I translate that, it's really to not just go manually, add on intelligent methods, make sure that the organization is higher on volumes, sure, but smarter volumes. See that we can uh make sure that we have income stream coming in faster and updates, and we can handle that. Um, and I think if we increase complexity, we also increase friction between people. Yes. So that's where we have the change really being tricky.

Henrik Göthberg: 15:44

But okay, so but let's before we move on. I think we are doing this is the foundational thesis for the rest of the conversation now. Another way of talking about exactly what you said now, and I'm I'm I'm I'm gonna try to be more practical and exemplify stuff. So let's say we do a use case and we build a point solution around that use case. Yeah, and then that works, you know, one workflow uh let's say for sales, and then we build it in in as a point solution, and then we are we are great. We are we are taking that point solution out to all the different sales teams all over the world. Fantastic. Then all of a sudden now now we have another use case and it and it's something for uh finance. So we do another point solution. And and what happens now is like worst case scenario, every new you're starting with a pilot in uh for sales in one market, and then you do a point solution here, then you do something slightly different for the next market in sales. So exponentially now, if you think about it, each new market becomes a little bit of a point solution, yeah. And then you add the point solution spaghetti in in another dimension. So all of a sudden now you have actually you have exponential total cost of ownership. You probably have an a complexity. I mean, like they use the sheer way of adding to this becomes sort of trickier. Yeah, so this is this is the point solution problem, you know, that you get an exponent it total cost of ownership of operating this complexity goes out of hand. So then so this is one failure mode. Another failure mode, I would say, oh, we have this old old old system, and now we need to start from scratch with a massive data platform investment, and we we we we put a lot of money into platforms, but no one is using them. So it's new stranded technology. So so when we are talking about scaling, for me, this is some sort of golden path. How are we use case-driven, finding it, yeah, but not ending up in point solutions spaghetti? So we need to be business-led, yes, but we at the same time we need to be be industrialized in how we take things. I mean, like that's sort of summary. Yeah. Do you buy into that or do would would you frame it differently?

Helena Hörnebrant: 17:57

Well, if I hear you, I mean, how do we really scale with with um well? Let's start. I think we all agree that we need a value, right? We need a value. So it doesn't actually have to be business. I don't care if it's business or IT driving, but I I really care that it's a business value. Okay, value.

Henrik Göthberg: 18:17

Good.

Helena Hörnebrant: 18:18

And I also think that we can we should definitely build local use cases. And we have done quite a number of cool use cases, but it it's really fast become like a pilot perjuring. We're just so happy about our sensational use cases, but we we are no ability to scale it. So what we need to do is really to have a cross-functional team where some of them focus on the local use case, some of the team members focus on really to get the value thesis in so they can adopt it. And then some of them are really looking at the from the platform perspective. How can I uh create reusable components uh that we can use for the next and the next and the next? And I think that part is the number one. So we need to establish almost like two operating models: one that focuses on the use local use case, make that impact as great as possible for the local market in our case. And then have one that owns the platform perspective of reusability, but then also take responsibility for enterprise capitalization, if you will. So have somebody that makes sure that the next market and the next market and the next market is using it, possibly a bit personalized. But we can't ask that one market should think about seven markets or eight markets. We we we need to find a way that we have people that's smart enough to know what the other seven markets need. So I think maybe two operating models at once, or even three. If you're gonna have a long platform story where we change the big

Henrik Göthberg: 20:00

Platforms, but the the way I see this now, let's see. I'm almost seeing a graph in my head where on the on the y-axis, we are talking about value creation, the use case, the workflow we are fixing. We need to go there and really fix real hardcore workflows in the business, which ultimately you go all the way out to one market unit and their particular workflows and fix that. The use case view. So this to me is the y-axis then becomes value creation. The other X axis, the other so the other model, so to speak, is that as we go and we are asking ourselves, now we have done this use case, how do we need to decompose or refactor that in order for this to industrially scale? That becomes the value capitalization, harvest and reuse storytelling. So what you're saying when you're saying there is one operating model or there's one part, one part of the operating model at least that works towards the use case to find the real friction.

Helena Hörnebrant: 21:03

And very short run, maybe 90 or 120 days. It should go really fast, really uh this is this is to prove the piloting operation.

Henrik Göthberg: 21:12

Yeah, and then and then and then it becomes how do we scale this up, which is then the capitalization story.

Helena Hörnebrant: 21:17

Yeah.

Henrik Göthberg: 21:18

Which because it means then we have need to have people and practices and ways of working to find the right use cases and to try them out and then innovation lab to make them fast into to show what they're all about. Yeah, and then you need to have a way to continue doing those at the same time taking it home to the horse, you know, horizontal view. I mean, like this is the vertical view of the value creation versus the horizontal view of investing in scaling and industrialization.

Helena Hörnebrant: 21:45

Yeah, absolutely. But I uh yes, absolutely. And and on that topic, I think also we need to consider how we find the use case families because if otherwise it becomes just the smart ideas flying around, people feel extremely creative, let's do it, yeah. But we want to have what should the treasury department really deliver? What can we help them to deliver? That might be one or two or five use cases. Or what does the sales need to deliver? What do they need to make better point sales with the customer, or even increase the customer experience? So we need to really attach our strategy or the way that we uh approach our customer to these value streams. Otherwise, we become just a use case factory.

Henrik Göthberg: 22:32

Okay, so you you said one keyword here really fast. You said you used the word use case families.

Helena Hörnebrant: 22:39

Yeah, or as Scania would probably use as a value stream or yeah, so so let's see if I get it right.

Henrik Göthberg: 22:48

I mean, like I remember back in the day I was uh working um coming in uh after a large management consultancy, and they said, Oh, we have this use case, uh it's called profitability. Oh my god, that that is that is potentially a use case theme or a use case family because the way we can bring it down into smaller, smaller pieces is so many different ways you can work with profitability. Of course. So is that what you're saying? So you need to find what are the fundamental areas or levers we want to pull in order to I mean, like to have thousands or small product teams for each use case doesn't make any sense. But if we can group them in families or themes, that's what you were saying.

Helena Hörnebrant: 23:33

Yeah, with really leading KPIs, yes, how we want to measure that company, and then rather have uh OKRs or sub-OQRs to really measure how to get there, and that's where we should find the use cases.

Henrik Göthberg: 23:47

So the families, I mean, and oh we can have a different conversation. On I I started to stop, I started to stop using the word use cases.

Helena Hörnebrant: 23:58

Now you make me happy.

Henrik Göthberg: 24:01

Because it it it detracts from the fundamentals of the core workflows in an organization where we want to create value, and use cases easily becomes a tech responsibility when a when the real value lever sits and you know with the value capture sits always with the business as well, of course.

Helena Hörnebrant: 24:17

Of course. So I think so. We have to start from there, and I and I think otherwise it's just it becomes just a big, huge backlog. People pick as they please, whatever makes them tick that day. And we need to really make sure that we are so customer focused that we don't do anything that doesn't bring value to the customer.

Henrik Göthberg: 24:37

But now you now you're coming into data's core territory to think about portfolio modeling and portfolio organization, which we which essentially then how do we create value lineage for you know for what value if I do this and this and this in the back end, it leads into what values, and and in the same way, then what cost? So cost and value lineage is very, very hard to do if you have just a plethora of point uh line items. Yeah, yeah, yeah. So 200 line items, how can I create a lineage between you know platform investments and themes and stuff like that?

Helena Hörnebrant: 25:14

And I think a lot of tech people get so happy about the use case and the tech that they use to get there that adoption becomes very um very uh narrowed focused. Uh, and I would really have adoption take equally long time as almost the use case uh uh development of it. So scaling and concurrently.

Henrik Göthberg: 25:38

Yeah, so wrapping up what we talked about scaling that is value driven needs to be some sort of sorted in families in order to have a portfolio view is one key area, and then we have the other key dimension of uh bimodal in terms of productizing value creation, the use case or the pilot workflow versus the scaling value capitalization. So these are some key words we are circling around now.

Helena Hörnebrant: 26:03

But I think uh we we have possibly forgotten one important possible. Probably more, probably more. Keep going now. You're being humble. Uh but I think governance, uh, we we we in big organizations like ours, we we will continue working traditional, and but we need to slowly move over to the new ways of working of cross-functional teams, and that uh governance, we need to be supportive of uh reducing the time to market. We need to be supportive to use the money for the right topics and the right value. Uh it's so easy that we become controlling. You prove to me your point instead of the other way around. So we need to see how we can set up a governance that's supportive for speed, quality, and value.

Henrik Göthberg: 26:54

So, are you are you in a way now putting in different you're you're entering in the topic of scaling? I'm reading between the lines, so maybe I'm reading it wrong. That that we need to think carefully about what has been the way to govern and manage scaling in the past and slightly now, other mechanisms or other ways of thinking about this is sort of you know, this scaling topic, is it changing, you know? You know, is it is it is it something we need to think about differently moving into 2026 and moving forward? Is this high on the agenda for everyone? I think it is, but is it different to how we thought about this? Um what were you trying to say here with the cross-functional and the governance? Yeah, don't we have that?

Helena Hörnebrant: 27:37

I mean, no, I think uh I think we have it. Some places we have it, and in some places we don't, but I think with AI and new emerging technology coming in so fast, we need to realize it's not a tech project as per se any longer, it's a business uh or a company project, and with that company project comes uh cross-disciplinary uh people sitting together to solve a problem. And we need to really make sure that they can work autonomous autonomously, and if so, they need a structure with a cadence that's right for them, but also with people that can make decisions the whole uh 1 to 120 days or whatever it is that it takes us to develop something, and I think uh systematic, disciplinary, but also mandate-wise.

Henrik Göthberg: 28:30

But okay, I hear in between the lines now that maybe what we were looked at as scaling 10, 15, 20 years ago, you know, when I went back to university, we talked about economies of scale, right? And the whole idea around scaling was very much connected to efficiency. And we worked our separation of you know, division of labor. We become more efficient because we put all the business controllers over here and we put all the tech people over here. And all of a sudden, now with the age of AI, or actually, we we are actually talking about what scaling means in a in a VUCA context, you know, volatile, uncertain, complex, uh ambiguous, and we are talking about scaling in a context where the invention rates of new technology we could use is something completely different. So to scale and be efficient now, we need cross-disciplinary stuff because we need someone with the tech expertise and the business expertise. Yeah, so I'm so I'm I'm I'm sort of seeing in between the lines that scaling the way we went about and thought about it 15 years ago is slightly different to what you are sort of hinting at here.

Helena Hörnebrant: 29:49

I think so. The speed that you're mentioning. I also think that business is more distributed, yeah. We don't have time to uh fix everything ourselves, we don't have time to focus, but also we're not good at everything. So we need partnership, we need to work in ecosystems, and and that distribution of um um how we perform or our operating models is both key for cost, but is also completely key for innovation. On top of that, our margins are getting reduced everywhere, uh, and and I think we need to have a new way of thinking about producing more on less resources. I'm not just talking about people, but and not just doing more on fewer people and resources, but also smarter. And those two in the VUCA world with speed of technology development puts us into something completely different.

Henrik Göthberg: 30:48

Okay, so let's summarize because you said now we have a productivity frontier, is an academic word where we sort of it moves forward all the time, and and to some degree you need to be at the productivity frontier to be competitive, yeah, and you can be a little bit behind it, and maybe it's in and the productivity frontier now. We can see is different in different industries in terms of how effective we are, how smartly you use data. This is one. So the productivity frontier is always going up. Yeah, that's that's how it works.

Helena Hörnebrant: 31:21

And the expectations of producing more on less is there the whole time.

Henrik Göthberg: 31:27

Oh and and do you need to be somewhere in balance with your industry on the productivity frontier? Then you had another concept next to this now, which is the innovation or invention pace. So the invention rates we see right now, especially in tech like AI and data, where we're talking about doubling the capability of the AI in long tasks and stuff like that every you know seven to twelve months. That invention rate, the difference of what could be done versus the way we are working at home, it be creates a growing gap. Yeah, and that gap to me becomes a threat surface. If you're not closing that gap, if you're too far away from the art of possible, someone you know that someone else will go in there, and that's the new productivity frontier in your industry. Yeah, so so then you have invention rate and you have productivity frontier, and then on top of that, we say BUCA, which is meaning geopolitically and you need to be resilient and we need to be resilient and super flexible. So we need to do scaling from an adaptability point of view. We need to do scaling from learning and adapting new tech fast. We need to do scaling from the point of view that we are questioning the productivity frontier. And the difference is that maybe we could 20 years would define an assist and be quite stable, you know, draw process lines, and then we work like this, and that would hold for five years or 10 years, and then we do a transition or transformation. Yeah, now it's something that kind of moves the whole time, the whole time, and then what is scaling?

Helena Hörnebrant: 33:08

Yeah, yeah. So the linear scaling is basically gone, I would say. Very few uh can do linear scaling on the long run. But I and uh then it puts more pressure on us on how to innovate as well. And if we uh put the how to innovate, we need to be much more with optionality, we want personalization, the customer has greater expectations. Uh and uh and uh with that uh we we need to be able to adapt our products and services offering service offerings to the customer more, and and that's not a chance in the world that you can be alone in that journey. No, no, so the ecosystem innovation is gonna just grow and become bigger and bigger. And for us as a financial services, that can also be a product. It can be somebody else coming in and helping us. It can be um somebody has helped this customer, and now would you like to buy that? So we we need to both buy technology, we need to buy innovation, and we need to buy partnerships that actually do parts of the customer journey for us.

Henrik Göthberg: 34:18

It reminds me of um we had a the guy uh Hugin from Riksbanken. He he basically was standing on the barricades. I hate when we say buy make or buy, the old procurement story. Yeah, he said it like this everything today is buy to build. And what he meant with that is very closely related to tech, that we always we all start somewhere buying something. Um like you don't go out and build and build your chip, even if you go and get buy a GPU hardware, you buy something, and you need to design where in that stack should I buy, yeah, then to build your product, your process, your workflow. Yeah, and and the whole what he was so strong about that, and I love him for it because his true is it's so true, the old make or buy logic came about from procurement saying we shouldn't do our own software, we should build a we should buy a business application. That logic that doesn't really hold anymore in this distributed ecosystem world and where we have data and we have AI, and where we're definitely not gonna do everything from scratch. But even if we buy the perfect, you know, uh vendor, what are you gonna build with it anyway? So it's a different mindset, it's a different mindset, it becomes the modular modular, but also it says something on what you said now how we build our industry ecosystem and partnerships by to build is I can take it all the way into an insurance product. Yeah, and we should build the brand, or we should build the meta product for this insurance product, but I'm not gonna but underlying here, I'm gonna use Santander's core, you know, and then you can take that story to data, yeah. The whole what buy by to build or partner to build, I think is such an important uh lesson in this uh game.

Helena Hörnebrant: 36:17

Yeah, exactly. And I think we even used to say that whatever is business critical or customer critical, we would do ourselves, yeah. But we don't have that time exactly.

Henrik Göthberg: 36:30

So even that, even that is not that simple anymore.

Helena Hörnebrant: 36:32

It's not that simple anymore. So we need to find partners that uh might not have, for example, we have a fantastic distribution network uh that we can actually lure in, or we can uh offer one thing that they can't, or they can offer one thing. So we also talk then about different business models. How do we share that?

Henrik Göthberg: 36:53

Yeah, but and and and I think you I can also flip it. Buy to build also needs to be understood. If you say make or buy, oh we we now we bought this fancy tech. So now we don't need to have engineers. No, oh my god. So now even whatever you're building, you need to still build something, right? So so, but that then is what is that something that is our IP or that is our flavor? Yeah. From how we do data products to how we do AI services to how we do governance to how we do financial instruments, how we do product. So it it this build this goes into industrialization and scale. Don't think this is nothing you do alone anymore.

Helena Hörnebrant: 37:38

You're not alone any longer at all. This is an ecosystem, and I think that ecosystem uh scaling. So if we just I would say even five or ten years ago, we we did organizational scaling. That was the thing. We added people more, we might have in-sourced people in in um brackets that were a lot lower uh salary brackets than we had, etc. But that's not the point any longer. We need to scale everything, including your ecosystem. And I think companies um that scale internally with headcount and assets and capabilities, it's superb, but they can only get to a certain amount of speed and scale. Um, so I think uh platforms, data sharing, open standards, um really leveraging the network you have and see what you can do that's unique instead.

Henrik Göthberg: 38:34

Yeah, and and then then you the the whole network effects. You can you can talk about this on commercial ground, but in the end it goes all the way to how do I share the data between two different partners? Hardcore, yeah.

Helena Hörnebrant: 38:47

Hardcore, safe and sound, safe and governed and all this.

Henrik Göthberg: 38:51

Before we jump into it, there is of course one angle we need to put into this. When we are now talking about scaling, we've already gone for it for about 40 minutes. Is it fundamentally different how to think about that in a large enterprise versus small medium enterprise, or you know, the startup um startup hyperscaling journey? So are there any caveat, or are there any distinctions you would like, you know, when when we are talking about scaling now, and and we are of course talking, we are starting this from an enterprise point of view, right? But do you think this end is worth noting something, or is it the same in different ways, or how do you see that?

Helena Hörnebrant: 39:31

So I think it's a major difference between a large enterprise and a mid-sized or a small company. Yeah, if we compare uh a big enterprise have a very complex system to run, whereas a small one or a mid size have a very adaptive movement, they can be bigger here, smaller there. You know, it's not that big of a deal, uh, the system of it. Unless they're an ecosystem player, then they need to know their ecosystem. Uh in a big question. We need to rely more on processes and governance. Small companies also need governance, but they can actually, through relationship and energy, and engagement and positive reinforcement, really get people to move in the right direction. We in a big company, we focus a lot on alignment. And I know everybody's like, oh, there's a mountain alignment that we actually do is crazy. So, politician or not, we're doing alignment a lot of our time. Whereas in a mid-size or a smaller company, you focus more on engagement. And of course, in a big company, we need to focus on engagement as well, but there's not enough time for you to engage in everything. So you need to be more on an alignment basis. I think a big company or enterprise, you can't perform large scaling or changes unless you do it per design.

Henrik Göthberg: 40:57

Okay.

Helena Hörnebrant: 40:57

Whereas in a small one, you just change by doing, just learn, just do, do, do. In my setting that I have today, we have um 65 plus countries uh all around the world. We have a lot of legal entities. Um we are um maybe um Yeah, we didn't talk about that.

Henrik Göthberg: 41:18

No, can we can we say something about how big is the portfolio of transformation? Is that official or official?

Helena Hörnebrant: 41:25

Of course that's official. But it it's uh quite uninteresting.

Henrik Göthberg: 41:28

But we have we have a No, it says something about the scale we're talking about.

Helena Hörnebrant: 41:32

Yeah, we we have uh many millions.

Henrik Göthberg: 41:35

No, it's more than millions.

Helena Hörnebrant: 41:37

It's billions, of course. Uh no, but what I meant is that we can't talk about that, maybe of course we can talk about it, but I just mean that in this context, it's more interesting to know that we basically double and tripled in in the last uh number of years. We have um in that we have also uh had the business going on in very small and mid-sized cells out there, yeah. So using and they are really leading us, they are making the business.

Henrik Göthberg: 42:07

And they're small out there while we are big at home at the same time.

Helena Hörnebrant: 42:10

Yeah, so we have about 20 large uh companies, and the rest are a tail of uh 40 plus companies, so that's a really small one. And I and I think that uh by that we actually have both the large enterprise and the mid-sized and small company. So if I go out in the small company and I start to talk to them about only policies and processes, I drive them crazy. They want to do like what do we do now? How do we do it? Are you gonna be engaged? How much energy are you gonna put into this? We we can run this on our own, but we need your support.

Henrik Göthberg: 42:44

But this is I think this is so important to understand how the how to how to game this or play this or navigate this, whatever word we want to use. A central muscle, which is quite large, but in reality is very much an organization Hub and spoke in terms of you have a central approach, but then when you get out to the legal market in Peru, yeah, it it's a very fairly small office.

Helena Hörnebrant: 43:08

It's a very small office that we have at Financial Services, and um and I I must say that a lot of our uh people up there they're uh really heroic in so many senses. They do so much with uh uh being quite a slimmed organization, so they need to be good at many things.

Henrik Göthberg: 43:24

Whereas if you're a large corporate and you're on the head office or that, you you you need to be quite specialized in order to do your your scaling to take this traditional ERP style management consultant going into a lot a large American standardized bank and you just going the same for all, it's not the same ballgame here because okay, how many markets are you in? 60 right now? No, we're 65 plus. 65 plus now. So and and we're talking loan and lending and insurance.

Helena Hörnebrant: 43:53

Yes, we're talking that. We have other product, but the majority of our products is that, but we have some more sort of a bank or not?

Henrik Göthberg: 44:00

No, you don't you don't call yourself a bank.

Helena Hörnebrant: 44:02

But of course, in uh we have almost 70% uh of our flows are in regulated business uh units. We do have bank status in one or two of our companies, but we are a credit institute.

Henrik Göthberg: 44:15

You're a credit institute, but it's the interesting point. Like if you compare yourself to most, a lot of credit institute are strong, one in the Nordics, one in Sweden, and even really, really big ones don't operate on 60 65 markets.

Helena Hörnebrant: 44:29

No, we are one of the biggest in the world. In terms of footprint like that. Footprint like that. Uh and that, of course, is uh part of our great uh uh you know, thinkable, act local kind of business, and but the rest of the market is going the other way around. So when we are meeting our partners or ecosystem players, they want to have a centralized uh and they don't understand. I mean, my boss business model is built on very central companies, large size and portions. So they look at the enterprise level and then don't realize that we have a lot of small uh companies. So this is really the area where I've been most challenged in my life because you need to really balance this in a good way. And I think in that sense, if you look at um uh how we did uh uh our carve out in that, that means that there's a lot of companies we need to do. Carve out, it's a lot of entities we need to carve out, and we just added four uh new brands to our financial services. So that journey has been really laying out the foundation. Now the next step is of course, how can we optimize this? How can we move in the direction?

Henrik Göthberg: 45:44

But here you went quite fast on something which is from a scaling point of view quite interesting. So let me see if I you correct me if I misunderstood, but uh, as I understand it, the starting point here is Sconya Financial Services, where basically you had you're doing loan and lending and insurance for one brand. Yeah, and then how the other brands, even if they were owned by Volkswagen and Troton, how they did insurance or f or or lending, that was completely four different ways.

Helena Hörnebrant: 46:17

Yeah.

Henrik Göthberg: 46:18

And then basically even more than that. Even more than that, yeah, but but but then you basically legally shaped Troton Financial Services. I think you went operational 23 as as a legal entity, where basically now you have a strategy where you have a core center, this is the enterprise view, who are not only serving one brand, but four or five brands, and those footprints of those four or five brands who is in 65 markets is a little bit different, but you know, so you have you go to one country where you might serve four different brands.

Helena Hörnebrant: 46:54

Yeah, exactly.

Henrik Göthberg: 46:54

So now it's you exponentially yeah.

Helena Hörnebrant: 46:58

So the whole point was so we have been really successful Ascania Financial Services. The other brands have had also successful uh Scania Finan Services, but uh not to spread, and they've used partners or white labeling or other parts of it. Um, and we thought instead of going four different ways, if we go one way, we will be stronger and really cherry pick what's best in each of these uh financial services. And so in less than two years, we added uh on uh 20 entities, something, and four new brands, and uh we really did a huge transformation from a technology perspective, but also a legal perspective. So what was really chaotic in the beginning is now more and more taking shape. So uh now that we've come so far, and I must say that doing this much in two years or two and a half years is is at the same time as we're growing, was one of the toughest challenges that we had. And I think all of us leadered leaders in this had to reinvent a bit ourselves on how to do the steering and how to go about the problems we had every single day on that. Uh, but we are here to solve problems. So that for me is a positive thing. Um, and I think uh what what we were used to approposed scaling was that we design something and we go there and it takes a lot of years and all of that. And now we just forced everything over to get the platform straighten out, the legal platform, but also the technical capabilities of that. Now we need to take the next step. How do we scale to optimize the structure and grow an even smarter way than within the guardrails of eBIT and return on equity?

Henrik Göthberg: 48:54

Because now you get to a point, and if you're not taking this is the point. Now you've you've done some part of the scaling, but back to the the way we started to scale with decoupled OPEX to scale without adding complexity, that's the whole journey now where you need to sort of both now innovate with new products. In but in this new context, you need to be very, very careful to not add more complexity. So it's really an interesting game.

Helena Hörnebrant: 49:18

And we we we certainly did add complexity, especially on the technology side.

Henrik Göthberg: 49:23

Step one, yeah. Step one. No, you didn't add, but if you put four buckets into one big bucket, yeah, yeah, of course, it's uh when you look at it next to I mean the same complexity was used distributed before.

Helena Hörnebrant: 49:36

Yeah, exactly.

Henrik Göthberg: 49:37

So no one saw it, right? But then you pull it all into one bucket. Yeah, now you have some industrialization left to do.

Helena Hörnebrant: 49:43

Yeah, exactly. And and I think also uh we it's time to cherry pick what did we do good in one uh country, and even if we use Connie Financial Services base, for example, in Mexico, we used the international, they were way further ahead. So it's as a key as a market unit. Yeah, so there's other countries that's been much further ahead and more smarter than we in some markets. So so we we have learned, we have really, really put the learning on. And I think, but from long planning cycles to always do the next step and the next step and the rest, next step and the next step. And I think that's really the new way of thinking. And I think uh next decade of that is we're gonna have to experiment, and maybe experiment is a provocating word for some, uh, but at least test, learn, test, learn, analyze, test, learn, you know, that cycle and iteration will go the whole way through. And I think what we see when we don't do that smart with data, for example, we see that uh we can't reach and anything that normally felt, you know, I can just take up my computer and send an email or even book a room wasn't possible because we were on different tenants, we didn't share the same data and all of that. So that part we have done in certain other countries, but not all of the countries.

Henrik Göthberg: 51:08

All right. So the way we are the way the conversation is going now, I'm I'm I'm gonna give you uh I have two things I really want to go deeper on and uh discuss with you that I think is quite helpful, also for the listeners as a learning point. So I have one topic where I know since we've been talking about these topics over the years, you and me, in different ways, we've been fine-tuning our keywords or things that we think are super important that we want to distill into our organizations and our projects, how we work. So I call that sort of keywords and principles that are sort of when you're moving into scaling in the VUCA world, that has been the core problem. And now we, you know, what works and doesn't work. So these are the core principles or keywords that I would like to elaborate with you on because we I've stolen from you, I think I gave you a couple over the years, and so this is one angle to go down that route. The other route I think would be very helpful is to take a concrete example and try to think about what we've said so far. I see some sort of golden path between the value creation and the value capitalization. You call it bimodal. So my bimodal for me is the golden path to balance those two dimensions. So either we take an example or we'd look at some of your key words and principles. I think both of them are very helpful to sort of start pulling this down into more concrete levels. But which order should we take them, or which one do you feel for most of those two angles?

Helena Hörnebrant: 52:42

I think we need to start with some of the words so people can reflect on those.

Henrik Göthberg: 52:46

So then we bring them with us into example. Perfect. So okay, I will test you from this from memory a little bit. I love, I'm just gonna wrote them up. I love how we talk about pull and push. I use that a lot. Um change by design or systemic design. I call it systemic balance chain, change, blah, blah, blah, years back. If anyone look looks at my pod. So systemic design, uh, change by design. Then we've talked about discipline and perseverance, trust, and now lately I've you know uh agency and autonomy and those kind of words. Okay, so so this is some of your words, some my words. Okay, if you would do if we if you would summarize your kind of principles or key words, what is the list? Then this was my give you don't you don't need to take mine.

Helena Hörnebrant: 53:39

No, but I think that um first of all, I think we talked a lot about is systematization, yeah, and that's the corporate view of it, and the personal view of it is discipline.

Henrik Göthberg: 53:50

Okay, elaborate here first.

Helena Hörnebrant: 53:52

So systematization is really to see even if everything changes around us, even if we have a new target, is the constant moving target, we have a market that is changing all the time. We always gonna find common denominators that we can use the whole time and systemize through processes, maybe tools, or the way that we engage with each other, that type of systematization so we can trust the way that we are doing it.

Henrik Göthberg: 54:23

Let me try one. Years back, we had a conversation where I said, no, no, no, you need to build uh the engine. And you said, no, no, no, you need to build the muscle. And what we were talking about was the same idea of systemization around innovation to adoption and scaling. So so the world around us is shifting, everything, everything is coming, and all of a sudden we get we could do this, we could do this. So having a systemization of how we take invention and see how it fits use case, and then from use case systemized to platforms. This to me is systemization, both in the in the core principles how we scale, but also systemization in in innovation and design and change. Yes, is that what we're talking about?

Helena Hörnebrant: 55:07

I think so. And I think uh we we have often referred to as trust the process and just stick to that, but sometimes systematization allows for more elaborative processes, so you can be uh really large using the process uh more in the beginning or less in the end, or however you want to combine it. But systematization is really about how you go through it and where you normally um see where you get stuck, or where you normally find your best ways forward, or how you as a team solve it even better if you solve it in the morning and not in the end, or you put hats on, negative and positive hats, or whatever it is. So, systematization for me is a bit larger than the process. This is how you approach it. Yes.

Henrik Göthberg: 55:55

Okay, let me let me now try. Because we had this conversation with Matthias Frost at Nordia, and he put it like this. And I think we're talking about the same stuff with other words in the bank, Nordea in this case, we've done a lot of good work to figure out our customer journeys. We are quite quite work, but when it comes to the employer journey, or even more important, the innovation to adoption journey, yeah, we are it's really unclear. And and and and and now with AI, who should I talk to? What it's so much legal uncertainty. So here is what I mean with systemization. Yeah, you need to educate and create clarity on what's the way we approach things. Yeah, so when you say trust a process can be misleading because it's not we are not saying this one process that rules them all. Yeah, no, we have a process around innovation and governance, so we have a systemic way to bring new things in, yeah, we have a systemic way to flick on AI, we have a systemic way. So we're trusting that we are not so we are governing, we are it makes us smoother, but it also makes us more risk conscious and all that. So this to me is systemization, which is hard, right? It's really hard because we are muddling through on that and we're educating people on the tech, you know, blah blah blah. Yeah, but we're not educating, we don't have an educational clear view on systemization that we can teach.

Helena Hörnebrant: 57:20

No, I agree with you, and and that approach also, if you take it from a personal view, becomes a discipline to our leadership principles or our guidelines on how to approach a problem. And uh at Tartum Financial Services, we're using a lot of the principles all the day. We try to really break it down. And I think that becomes even more important when it's ever changing. And then we we come over to the change by design or continuously designing cool here now. If designers used to be some extra and cool uh for customer journeys and others, they now become central because it's gonna change even in the process of those 1920 days sometimes. So every time you come out on the other side, we have slightly changed the pictures. So the the big target picture is changing. And I think that continuous design and redesign, you have to get used to that and be curious and even happy about it. What's next? How is that gonna look like? Why is it changed?

Henrik Göthberg: 58:25

Okay, let me take my speed on what then so you said continuous design and continuous redesign. And you're you're um I can go into my rabbit holes of Derek's IP here, but it starts somewhere of the fundamental understanding that there is no process stable process anymore, they don't exist because as soon as you start drip feeding new data, new machine learning, new things, you are continuously tweaking the process. So it's continuously changing a little bit like how people work or what we do is at that pace. And that sort of that whole field, I even had that on Data Innovation Summit in 2020. We had that mantra, process becomes product. And what we mean with that, we need to approach our core workflows as they were with product thinking of continuous design. Like no one in the right mind building software today has one product and do a release every three years. No, that's the you know, we do we do continuous devils, we do CICT. Yeah, they we used to release new things all the time. Yeah, and we need to actually get our process view and our organizational view to live in that speed, in that cadence.

Helena Hörnebrant: 59:39

I agree. And what we call continuous improvement is a superb word, but we used to take only small, small, small, small steps. Now those steps have to be bigger, and you have to actually like them bigger.

Henrik Göthberg: 59:52

Yeah, and and I I'm not sure it's correct, but I've been in the when I've been trying to distinguish what do we mean. With continuous design compared to continuous improvement. I also talked about perpetual in five, six years ago, I used the word perpetual innovational stage, where continuous improvement for me frames it, you know, we we need to innovate within the game we know. We need to innovate within the process we have. And we are doing lean, but the frame is quite the process we have. And all of a sudden now in the ecosystem world, all bets are off. You can innovate in every single direction. So you need to take a more continuous improvement for me belongs in the process world. Continuous design, perpetual innovation is the product thinking. Exactly. And they'll and why do we say product thinking? Okay, so continuous design now becomes really, really tricky if you have a long, very long chain and very hard-coded things. So here now we start modularizing in order to everybody as a cell can have a product thinking with clear APIs, so to speak, to each other. Yeah, then they can do continuous innovation. Because if you have that very, very long, hard-coded process line in ERPs and whatever, then it's tricky to do continuous innovation because as soon as you pull in the one end, uh the uh the knock-born effects is massive, right?

Helena Hörnebrant: 1:01:22

It's massive. So here I think we come back to the next word, and that's trust. And what we used the trust uh before, it's like, oh, we have one cool strategy, everybody knows it. Now let's run to that. But today that just is more of an ambition and a North Star. And then you need to trust the people you have around you, that they help to redesign and design, um, which means that you need to trust the people that they are competent enough to make it in the speed that we're doing, and trust the way that we are working, the common denominator, because you don't know the result all the time.

Henrik Göthberg: 1:02:00

Okay, this is huge because now we're talking about trust from a dimension of in order to go into someone and say the way we have done it can be done better, requires psychological safety within the team and with your interfaces. That I'm not I'm not banging on you. No, I'm just wanting us to do it better together. Yeah, so then you have the fundamental trust and what you saw high to the ceiling, hefti talks, we say in Sweden, in order to look at things that could be better without people getting defensive around stuff.

Helena Hörnebrant: 1:02:33

Exactly. It's not personal, it's not personal, it's just it requires completely different to go in uh exponential pace than to go in a lot of better pace.

Henrik Göthberg: 1:02:44

Yeah, okay, so that that that is super clear. Now I now let's do let's do trust and discipline together. Because what happens now, if we are working in a cellular fashion, the the major risk we're doing that is complete anarchy and fragmentation and siloed behavior in this. Whatever becomes these network effects, yeah, is all about how we are trusting and disciplined in how we treat the system and the and the and and the interfaces. So you need to have extreme disciplined thinking on system thinking, and you need to trust that we are federating this together, yeah, we are co-creating this together. As soon as the domain, if I use that word, thinks that the central platform team is shit, and they are we can do it on our own, the whole thing fails falls apart. The discipline to follow the patterns that allow data sharing as an example, the trust that you do your part, I I do my part. Yeah, if that's not there, you don't you don't get an intelligent mesh.

Helena Hörnebrant: 1:03:50

No, you don't, you don't get the intelligent mesh. You get architects and designers and facilitators become one of the most important roles that we have.

Henrik Göthberg: 1:03:59

But but but then leadership in terms of what you said, discipline. If everybody's allowed to use their own data protocols, you know, because they they can be a little bit smarter over here, yeah. And we but we had those principles and we were supposed to trust the central team to do their part and we can do our part. Yeah, if the trust and discipline falls out of the door, the modular approach falls out the door.

Helena Hörnebrant: 1:04:24

I know. So the leadership today is really about marathon, yeah, and it's ultra marathon, yeah. And we need to really make sure that people walk that walk together and have the talk together. The North Star keeps us from uh having results that's fragmented. Yeah, so if you facilitate your North Star and break that down, uh you you shouldn't get fragmentation, you should actually get a common uh journey.

Henrik Göthberg: 1:04:57

So systemic change by design or continuous redesign, trust, discipline, you cannot do modular without those things, in my opinion.

Helena Hörnebrant: 1:05:10

No, you can't. Uh, but I also think what we what we have discussed in between all of this without putting words to it, is that if we are now scaling in an ecosystem of cells, we need to make sure that those cells can work autonomously.

Henrik Göthberg: 1:05:30

Okay, that's the next keyword here, right?

Helena Hörnebrant: 1:05:32

And if we do that, that will provide not only speed but intelligence within that cell. So we need to follow the pattern on how connect the cells, APIs or data sharing, or uh even cybersecurity, how we share things. Uh so we need to really provide stability in that. We do that with architects and designers and others and facilitators and North Stars, but we need to be disciplined to work as a team and follow that. So we get the big brains on it, then we get intelligence and speed. And I think we can only do that if we have enough mastery in that group.

Henrik Göthberg: 1:06:16

Okay, so so now we're going into those things. We want to work modular, but to work modular, you need to now add the word autonomy. Now we need to break apart that word autonomy so we don't end up in fragmented anarchy. And we need to look at what creates speed and autonomy within that. So let's take that slowly. So the one you added it trust, and then you went to autonomy now. My first angle on autonomy is that it's one of those words that are misunderstood. Yeah. And basically, people are misusing autonomy for sovereign sovereignty or so they say, Oh, I'm autonomous, I have autonomy, but they want sovereignty. A simple way of understanding autonomy is I think your child in school is an autonomous student in that in that classroom. But autonomy means he needs to follow protocol and needs to act and behave in a certain way. This is number one. So you so you're always autonomy in a context that where you need to align with others. So the word autonomy, I know Spotify talked about how we need to have aligned autonomy. Aligned autonomy is actually surplus for the word autonomy means something already.

Helena Hörnebrant: 1:07:34

Yeah, I agree.

Henrik Göthberg: 1:07:35

If you go disaligned already, is the word autonomy in infers you cannot be autonomy without the context that you're autonomous in. Okay. But fine-tune, okay, aligned autonomy. The next dimension of autonomy, how you get autonomy that works, is these feedback loops. So how so if you are are we working autonomously, but you are not feed? I mean, like so when you're in a classroom, yeah, you get feedback from your students, friends, and your and your and your teacher, how am I behaving? And the same ways in a system, you know, in in in Troton or whatever, am I working more efficiently? Is my work in that part of the chain making the other guys more efficient? And this is sort of then feedback loops that needs to go the chain of you know down to things that you need to take care of, but also feedback up uh or somewhere it doesn't work for me. Yeah, so autonomy is fundamentally built with feedback loops, yeah. And feedback loops in the age of AI is encoded data, so it's not so it starts with feedback between people. So then we have the team humanistic, you know. I can say scania works not from its structural capital, but from its human capital, then I'm referring to these feedback loops really work well in Scania humanistically. Yeah, that story needs to be translated into code and data.

Helena Hörnebrant: 1:08:56

I agree. And in and even if you take it so far as you go to one of our market, feedback loop can actually be transparent reporting.

Henrik Göthberg: 1:09:06

Yeah, exactly.

Helena Hörnebrant: 1:09:07

And I think that what we are actually inevitably talking about is learning how to get to where we want to go.

Henrik Göthberg: 1:09:14

Yeah.

Speaker 2: 1:09:15

And then working with uh access of knowledge, working with learnings, adjust, and for people to do that, they need psychological safety and they need engineering cultures that improvements are really good. Deviation, love deviations.

Henrik Göthberg: 1:09:34

But you said something now with autonomy, and I I I went down the rabbit hole of feedback loops, but then you said another key word. In order to have autonomy within a team to move with speed, you also need the mastery. Could you elaborate on that?

Helena Hörnebrant: 1:09:49

Yeah, I think uh mastery is the urge to get better of something that matters. So uh if if you are a cross-functional team, you need to bring in the mastery of your specific topic, but in the structure of uh what you're creating. So if you're creating within a small cell, all people, whether you're three or five or eight, you should know the end-to-end flow. And you should know exactly one of your perspectives even more than the others, but you should know how they interact with all the perspectives. So it's mastery of flow and mastery of your domain.

Henrik Göthberg: 1:10:33

Okay, so profound. I think let me let me riff on that and see if we end up on the same play. I riff on it because for me, you said one way how you can talk about this and find if you want to read more on this, is about being T-shaped.

Helena Hörnebrant: 1:10:48

Yeah.

Henrik Göthberg: 1:10:49

So T-shaped means that we have a coherent understanding and communication within a team for flow, team flow.

Helena Hörnebrant: 1:10:57

Yeah.

Henrik Göthberg: 1:10:58

But then because we are in such a more complex world now, where someone needs to manage and know the domain competence, someone needs to know the data management part, and now we're gonna put an AI agent in here. So whatever we have now is at the smallest quanta, in my opinion, a cross-disciplinary team. And therefore, the team needs to have mastery. And to order to have mastery, we have to have a mastery of the common sense flow that the discipline here. Yeah, the psychological safety to work with people are not the same as me. And then I having the credibility or depth in one of the things I bring to the team.

Helena Hörnebrant: 1:11:37

Yeah.

Henrik Göthberg: 1:11:37

Are we talking about the same thing?

Helena Hörnebrant: 1:11:39

Yeah, yeah, yeah. We are talking about the same thing. So I think uh uh if we take it, because the words, of course, neither you nor me has made up.

Henrik Göthberg: 1:11:47

I mean, it's Daniel Pink, yeah, for instance, right?

Helena Hörnebrant: 1:11:50

Yeah, so the Harvard professor. So I think, but I think we use different words even before we came to these words. Uh but I think what he found out is that if you have purpose or North Star, you have a group of people that you work with autonomously, and you have a mastery that you contribute with both as a group and to unflow and yourself, you actually are in your best um uh position to deliver as a high-performing team. And you bring value to yourself and you bring value to the team. And I think that that's something that we need to work because if we're gonna scale what used to be purpose before I put this in, this is what we're creating. And now we're saying whatever you are creating, it's gonna change and change and change and change and change. So your purpose can't be that. Your purpose needs to be belonging, needs to be learning about new things, it needs to be contribution rate.

Henrik Göthberg: 1:12:50

I fully agree 100%. And now here's here's here's one of the don't agree all the time.

Helena Hörnebrant: 1:12:55

Like you have to actually that's the problem.

Henrik Göthberg: 1:12:57

It's it's the problem when we have been riff, we have been riffing on this for years, uh, you know, trying to make sense of normally we we don't agree on that. No, we don't agree, but but trying to make sense of the world. So this is a little bit like one-sided, I agree. What we have done that you haven't done, Mikel Klingwan, and we we tested these ideas in terms of design principles for agentic AI, and we think it's the same. So if because if you think about it, if you're now gonna have a team and you're gonna have mastery on the team, and you start taking sort of the coworker metaphor on the AI, how to think about the boundaries, the frames, what is the feedback I need to give to the team, or is the feedback I need to give to the data? We think those fundamental principles around modular scaling that we've been talking about now, you can you can actually deduce them now down to uh engineering principles for Igenic AI. You would probably use slightly different words to talk with the text, but if I look at it, what's the pattern? What are the patterns here? I think it's the same. What do you think about that? Have you thought about it? Like what we are thinking now, if we're gonna use machines as part of our team structures, it's gonna be very challenging when we start adding machine flows that is not aligned with the mandates and the views of the team for people.

Helena Hörnebrant: 1:14:26

We're gonna have to align them, otherwise it doesn't work.

Henrik Göthberg: 1:14:29

But I think uh Yeah, that was a left curve ball. I just want to try that out on you.

Helena Hörnebrant: 1:14:34

Oh, I think we we we need to work with that as we work with everything else. And I think um it's basically to go from linear talent growth, yeah, and then take that to exponential productivity per person. And then we're gonna have our digital colleagues. Yeah, and the digital colleagues in the beginning will be as flawed as us and as pre prejudice, uh, show prejudice as anyone else does. And we need to steer and put uh common denominators and governance on this in order for that to flow, just like we do with people. Is that what you meant?

Henrik Göthberg: 1:15:12

I think so. Because we've been trying to wrap our head around this because when you're designing an agent, technically you can dream of the world of what they should are. They could do this, they could do this, they could they could do this. And all of a sudden, now we build something that doesn't align with the decision mandates of the organization. Theoretically, you know, someone would tell me the argument is of course, why would you limit the AI speed and productivity to the limitations of the of the AI of the human limitations? I think it's a matter of safety that you need to go that way in order to then understand how do I redraw decision mandates of teams and humans, but but to use jump bit without without being in alignment, I think it's super scary. So I think it's a I think it's a stepping stone to emulate how we are working mandates, teams like that. I think then then when that is one and together, then you can reinvent with the machine and the man team as one.

Helena Hörnebrant: 1:16:17

But yeah, I think one you can probably use it differently. I mean, you can use it with with no limitations and then stuff them into the decision limitation that we have or put limits on it. Um I think what you're what you're referring to is ethical governance. Yes. How do we make sure that we align with the corporate's uh governance structure and also mandates? And of course, I can't build machines and all of the sudden have twice or three times my mandate. It would be lovely, but that would not be okay. So I think we we we need to put um we need to design work, governance, mandate, and human machine teaming with us and ethical governance on that. Yeah. And and then then it becomes what is ethical governance? That's something that needs to be an ongoing conversations within a company. Um and I think uh that that's gonna be a tricky one because it's gonna change all the time because it's gonna be more and more sophisticated. And unfortunately, since AI is built on normal human beings with uh very limited intellectual capacity and sometimes uh exponential capacity, we're gonna get get all of it.

Henrik Göthberg: 1:17:33

Yeah. I think break a little bit here because we need to think about the time. There's a couple of there's so much more I want to cover, and we have we can cover more of an more of the example. Now it was it some more words, by the way.

Helena Hörnebrant: 1:17:47

Was those two we went to all of the we went through all the world.

Henrik Göthberg: 1:17:51

Those are the key words.

Helena Hörnebrant: 1:17:52

We had uh systematization, yes, which is the the the personal side is discipline of that. Uh we were working with design and redesign, continuous designing. Yes, and we talked about trust, trust, and we also uh then we those are the words that I think you and I are discussed most, yes, long time, but then I know that all of us are we are constantly discussing what's the purpose, what's the purpose for the company and what's the purpose for purpose-led, yeah, but what is that? Yeah, because for some people they it needs to be personal purpose as well into this, and I think we talked about mastery and why that's important to get more uh reflections into the pictures and not stay too uh slim. And how do we make our teams autonomous to work with speed?

Henrik Göthberg: 1:18:40

I mean, like the only the only one here with where I think we'll had a deeper conversation is what is how to find purpose, right? And and uh one of the examples could be like, Oh, I I have a strong purpose, I own this process now. And this is for instance, like, is that a good purpose to have in the future when we are withdrawing maps all the time? Maybe not. And then we've had the conversation. Both you and me have a personality that we we love to innovate and drive change. But is that this is that a good deeper purpose or is that the personality trait that we love change?

Helena Hörnebrant: 1:19:19

I think so too.

Henrik Göthberg: 1:19:20

So we cannot we can we cannot sit still and now we get bored. Yeah, yeah. But I'm I'm thinking purpose. What is good purpose that is slow moving here now? That's I think it's a good question.

Helena Hörnebrant: 1:19:32

I think, first of all, why do we need purpose? So when everything changes around us, yeah, and I think purpose for most of us provides some kind of calm and stability. As long as I work towards this, I'm happy. We have purpose in our relationships, we have purpose in yeah, basically everything that we do. But uh well, I do a lot of things without purpose just for the fun of it. So maybe that's not true. Fun is purpose. Yeah, fun is purpose, right? And uh and then also align how we align ourselves across a team. That's a we bound we we are bonding based on that. Yes, please. But uh and I think also it filters the noise. Yes. Because I think when we don't have a common purpose or the right purpose for us, I think we often uh we become not stable within. We start to uh be a bit wobbly and we have harder time to make decisions and we also get too much noise in our head. And I think that's okay if you take responsibility over that, but it's easy to come into a team, and when you don't feel those things, you make the team wobbly. So you need to work with every single person in your small team, both what's the overall North Star, but also their purpose so they can be stable.

Henrik Göthberg: 1:21:04

Because I can take it all the way down to AI and having a clear AI objective function. What are we here to contribute with? And purpose to me highlights then what is the slow-moving dimension that we can now make decisions against when we are can potentially do anything with AI and technology. We could do this, we could do that, and we should listen to that vendor, we should listen to that. So it's really to cancel out the noise. So it's it's to get clarity, yeah, it's to get the zen. Why are we here? You know, and and then it has then you have the higher enterprise purpose. But I I we put it down to the core topic of um what is the core contribution of our team in relation to the rest. Yeah, and we we looked at that, me and Mikkel. If you design purpose that is too much inward looking, you can you can drive a purpose that makes you more efficient. Oh, we we'll we we drove down cost here, but you actually become uh worse and worse in contribution to your neighbor. So we were thinking very, very carefully on on on there there they're really the long-standing purpose for any spell is how do I boost the next? How do I boost the reacher? If everybody's thinking about it doesn't matter how much I cost or what I do, as long as I'm boosting the one using me more. So it's a little bit like finding purpose that is sort of customer-centric or even boosting or creating value-centric. Yeah. And and and that then means okay, I'm working in business control. So my I it's easier to say my purpose is accounting. How can you express business control from the point of boosting, you know, the business?

Helena Hörnebrant: 1:22:55

I think, first of all, if you you hear business control say they do accounting, they get very upset and like business control is something complicated. I know, I'm okay.

Henrik Göthberg: 1:23:05

But you see what I mean?

Helena Hörnebrant: 1:23:06

I mean, I see what you mean.

Henrik Göthberg: 1:23:07

I I don't know, purpose is important as hell.

Helena Hörnebrant: 1:23:10

I think it is important as hell, and I think it is in any relationship they're in, we should boost each other. And if you try to boost people the way I want to be boosted, you you're you won't get very far. So we need to make sure we know where your psychological safety and how you want to be uh triggered and boost, boost uh boosted, boosted, boosted.

Henrik Göthberg: 1:23:31

I don't know, but I I think it's it's simpler than that. That like if the mission as we as a team have, and what is or what makes us really if that is unclear, I think I think it's a slippery slope to how what is then efficiency, what should you prioritize, how should you build a team, what what competences are we missing?

Helena Hörnebrant: 1:23:50

I know, I I completely agree. And and what I think is uh when it goes that fast, you need to find a purpose sometimes without knowing the end picture. Yeah, exactly. And that's that's a tricky one. Yeah. And and I think why is it hard? Because if we are continuously designing and redesigning, that means we as a team need to reconfigure ourselves repeatedly. What are we gonna do? How are we gonna approach it?

Henrik Göthberg: 1:24:17

And then the purpose needs to be something else than the thing you just reconfigured. Because if you put your purpose in the thing that you're reconfiguring, then becomes mess.

Helena Hörnebrant: 1:24:26

You you become very uh messy on your identity of that.

Henrik Göthberg: 1:24:30

Okay, so now we covered purpose as well.

Helena Hörnebrant: 1:24:32

No, we need to move on.

Henrik Göthberg: 1:24:34

Sorry. We've this was a good one as well, but we because we didn't take that as far. I'm thinking we should either we go example now or we actually go more towards future outlook, because I think we we've been coming, we had the idea when we talked about this that what is becoming even more important in the age of AI, or you know, we I think we're going into principles that I think are quite leaning into VUCA and AI already. But I at the same time, I think the future outlook question, the way the way I framed it with you was like, okay, what we have said now, what we've talked about, we've sort of learned and and went this path for 10 years. Yeah, 10 years.

Helena Hörnebrant: 1:25:17

More, more, more.

Henrik Göthberg: 1:25:19

What what is what is fundamentally different the next 10 years? You know, so should we? I mean, like there's it would be nice to take an example, but at the same time, I really want you to elaborate on how you understand scaling in the age, I mean, like the future outlook of what we are doing, because I think some of the ideas of the different things we were exploring, they're quite good.

Helena Hörnebrant: 1:25:39

Yeah, I don't know how we're gonna do it, but uh that we have touched upon uh some of the things that's gonna change. You know, we go from efficiency to adaptability and optionality, or personalization, if you will. We go from process standardization to AI netative workflows.

Henrik Göthberg: 1:25:58

But I think all those things you can say it very fast, but but they they require their two, three-minute version. So this is one way of talking about. Also, like the way we had, I mean, like we can do the rapid fire on different perspectives of scaling product view, organizational view, platform view. Yeah, which one should we take? Which should we choose? I I'm game because we're not gonna be have time for no no no, so just take one.

Helena Hörnebrant: 1:26:24

I mean, if we look at the also the roles that Europe play versus Asia play and and uh versus um US.

Henrik Göthberg: 1:26:33

Yeah, we have to we okay. We need to spep place here scaling and and and talking a little bit more uh you know uh long term. Okay, let's do it like this. Let's see if we can do something. This is quite hard. I want to do a rapid fire. I'm gonna say a dimension, and how can we think about scaling in a certain perspective? So it's so the logic is this as a CIO or as a CDO, chief data office or AI office or or whatever, tech in relation to business and getting business on board. There's there are some very basic perspectives that you can um think about uh when we say scaling, hmm. Scaling, uh yeah, you know what I mean?

Helena Hörnebrant: 1:27:23

You're gonna try me out now.

Henrik Göthberg: 1:27:25

I'm gonna say it's I'm gonna I'm gonna shoot. So but you only have like it used to elevate the pitch. What's your what's your two cents, what's your main pointers? If I say to you, how should you think about scaling product or project? So basically, now we have one idea, we want to do a use case from idea, we want to do a workflow, and we want to test that idea, and then we what when it works, we want to scale that only one idea out large scale, 50 markets. So scaling products, scaling project, like what is your two things on the project?

Helena Hörnebrant: 1:28:00

I have a very smart employee who told me this, which I think is a good definition. If you're gonna scale projects, you actually throw people at tasks. But if you if you do scaling product, you you actually throw tasks at organized teams. And I think that says it all. Uh, people don't like to be thrown at things, they like to get a sophisticated task that they can solve complex problems and start to get better and better and better at it.

Henrik Göthberg: 1:28:31

So scaling product is literally the way to think about that, and the way to think about that is to have a team structure that can manage whatever product they're building. Yeah, that's simple. Yeah, okay, leave it. Scaling organization and at the same time, and and so there's the point when you scale organization, it's the chicken in the end. How do you scale up the budget or the portfolio in order to scale an organization? Or you know what comes first? I I know you need to go out and ask for some money in order to set an organization up, and then you need to, you know, scaling organization and scaling up the equivalent investment budget and portfolio. What's your two cents on that?

Helena Hörnebrant: 1:29:15

I think you need to break that down to me.

Henrik Göthberg: 1:29:18

Okay. I mean, like I I take the example now. You are starting in one end. This is where we should have done the case. Uh it let's say you have a uh I know we had one case lined up that we could have talked about, the ABS case. If I take that, we start we start in one end and we we are needing to scale that.

Helena Hörnebrant: 1:29:38

Asset back securities is what ABS stands for. It's a way to diverse our funding strategy.

Henrik Göthberg: 1:29:44

Because okay, let's do this rapid fire based on ABS. Now we're getting it.

Helena Hörnebrant: 1:29:48

Okay.

Henrik Göthberg: 1:29:49

So we have a product. The product is called ABS.

Helena Hörnebrant: 1:29:52

Yeah.

Henrik Göthberg: 1:29:52

And and and we need a team to scale it up. We don't want an ABS project, we want an ABS product. Two two two short minutes, what it's all about, and then we're gonna go through the other ones.

Helena Hörnebrant: 1:30:04

So we are diversifying our fundings like any other bank or financial institution or credit institution. And one way of doing that is to use a product called ABS, where we lend money on our portfolio.

Henrik Göthberg: 1:30:18

Yeah. Okay. So that's the that's the product. And the point has now been to do a pilot view to make that work. And when that works, and you have it centrally, now you need to scale that out.

Helena Hörnebrant: 1:30:29

Yeah.

Henrik Göthberg: 1:30:31

So we said first of all, product view. Now I go with that. What's the how do you scale the organization and how do you scale the budget equivalent rather than if if you know what's the work or what is the things to have in mind now to do that successfully? You've done this so many times. Yeah, exactly. So it's hard for people who've never done it.

Helena Hörnebrant: 1:30:51

It is hard. So in the beginning, we we we just need to do concurrent teams so they can work separately uh to capitalize on the product that we have created. So uh we need to scale absolutely uh the capitalization of the uh usage that is roll it out on more markets, get adoption rate.

Henrik Göthberg: 1:31:13

So here now, organization-wise, you need to scale out the muscle who can go out and enable an onboard market onboarding organization muscle.

Helena Hörnebrant: 1:31:22

Yeah, and then we continue to uh scale our organization for building a more sophisticated product at the same time. Okay, so that's sort of two operational models. One is to scale the product in itself and become more and more sophisticated, like you have on any app you get new apps and you update them and you get more feature and functions.

Henrik Göthberg: 1:31:43

So scaling in the use cases on the product itself, and then it's market geoscaling.

Helena Hörnebrant: 1:31:48

Yeah, and then of course we have scaling of people's uh uh skill set on how to adopt and take it in.

Henrik Göthberg: 1:31:56

Yes, okay. So then so that now you had scaling, so that that was number so scaling of organization, uh scaling of the investment, because what you are now building is uh an engine or a machinery or a staff who can now not only build a product but also take it out in the market, quite obvious.

Helena Hörnebrant: 1:32:14

Yeah, so my uh fabulous team that's doing this are scaling, so they have a of course a bucket for the product that they are allowed to decide on within their cross-functional team for that, and then with the the product owner of that, which is in our treasury department, and then we also have the capitalization and making sure that we adopt uh and uh uh onboard new markets, that type of bucket is per country.

Henrik Göthberg: 1:32:47

And how have you what is the what is the basic navigation or maneuvering or process to get to the funding that you need for each of those steps? You know, in what is your two cents because this is all about scaling, because when you do this, you now we need more team. Exactly. And and I find how should I put this? I think all this is more muddled through than really effective a lot of times in in in general.

Helena Hörnebrant: 1:33:15

I agree, we muddle through a bit on this, but let me be a clear, it's very simple.

Henrik Göthberg: 1:33:20

We're gonna scale in these three dimensions, we're gonna do this and this. Yeah, why is it modeled through and what should it be instead?

Helena Hörnebrant: 1:33:26

I think in the beginning you do model through to get everybody on board and to understand and to learn, and then you get your really talented controller, business controllers on board. That helps, that helps that helps and divides in the buckets and also uh make sure that the market knows what they need to pay for. So I think uh you need to be uh really good at uh controlling as well as uh creating product and adoption. Yeah, but you need to isolate it, that's the whole thing, and that's what we call bucket funding.

Henrik Göthberg: 1:33:59

What do you mean? Uh decompose it in into in is it clear? Yeah, in in in in what is the team structure that can run this?

Helena Hörnebrant: 1:34:06

Yeah, and you also need a specific budget for that uh part. So you don't mix it up with everybody else and everything else. You you really need to make them go autonomous to speed, and that's important to be able to do that.

Henrik Göthberg: 1:34:19

People are really muddling through on this, and they don't they don't know how to cut the budget, they don't know how to cut the portfolio. And unfortunately, I see more tech people trying to get through on this and they think they're gonna do it on their own when it's very simple. You can't. You need to work with your controller, you need to sort this out in buckets from the get-go.

Helena Hörnebrant: 1:34:37

Oh, yes. And I must say that when you get a phenomenal controller like ours, uh, who just helps us to set this up, it becomes a really simple Excel sheet where we all can run through and see and follow up. So we follow up really strictly true up.

Henrik Göthberg: 1:34:53

Okay. So now so and then you hinted on scaling out new ways of working and change behavior and practices because with ABS now, it actually allows us to do different things, but the treasure needs to work in a new way. Yeah. And the markets out there need to serve it, serve the treasury, you know, it's different for them as well.

Helena Hörnebrant: 1:35:13

It's different from them as well. So, how do you scale that out? So we we have two approaches that we do, and we like one more than the other. So when when we have cross-functional teams working uh on a speed and a budget and a quality level, we're very strict to have all three. Then uh they are helping us to write the requirements, yeah, check on the uh agile deliverables, and then test it in the end. And once they've done that, they know the process.

Henrik Göthberg: 1:35:43

So this is the systemic design, right? It's the systemic design. So they are the trust the process. We know it and we do it, and then we've done it once, and then that becomes a pattern, yeah, and the pattern is now scalable and repeatable.

Helena Hörnebrant: 1:35:54

The problem we have, or the challenge, I shouldn't call them problems, it's a challenge we have. We are so slim staffed out there, so sometimes we need to help the market, but as soon as we help too much, we don't have this skill set to actually adopt it. So, but we need to find a slimmer and faster way of how they can interact and test, not fully optimize it either, because that's never good either, but find ways so it becomes a dancing act, really.

Henrik Göthberg: 1:36:25

Uh co-creation together.

Helena Hörnebrant: 1:36:26

Co-creation, constant feedback loop, and making sure that also uh my tech people understand what they need.

Henrik Göthberg: 1:36:33

Yeah, but I thesis. You need then some in this organizational structure, you need the core platform or product team that really builds the core platform, and you need something else in terms of enablement who will dance with these guys, not to do their job, but in order to hold their hands until it sits and works.

Helena Hörnebrant: 1:36:55

Yeah, I think you're spot on. You you cannot do their work. No, so you need to have whoever wants to do this process needs to own it from the local market. It cannot be ownership anywhere else. But that means that we need to support that person so they reuse everything that we have done that's smart. And why do we want them to do that? Because we want them to add new smartness to the table, not invent what we already have. So if you you keep giving them a good platform, they can add on smartness to it and not focus on the actual foundation of it.

Henrik Göthberg: 1:37:32

Yeah, I like it. Okay, scaling out new ways of working, enablement, boom, bab, bold. Scaling platforms and technologies, you know, you know, one, I mean like you you you have such an interesting profile because on the one hand side you see how to scale out ERP, and on the other side, you now see how to scale out data, data product scaling tick. So, you know, so what's your two sense? You know, I think it's really important to differentiate the two. Oh, yeah. And you know, and then talking a little bit more about the data. But but could you use differentiate them? Because I get so trust when I don't have mature people that knows the difference and trying to do an ERP style rollout on data and AI.

Helena Hörnebrant: 1:38:18

Yeah, I think that ERP is a lot about standardized processes. Um when you uh roll out and scale up data products, it's more about how you can have standardized data products that can be shareable immediately, or patterns, or components, or AI or ML components for that matter. Um, and I think governance, compliance, um uh that type uh if it becomes Lego pieces, and you can put it together every single time the way that you need it in the speed that you need.

Henrik Göthberg: 1:38:56

And and I think this is maybe it's that simple that as long as you work with ERP, you're standardizing on process level and you're trying to hard code it into a standardized process, and and that's how it should work.

Helena Hörnebrant: 1:39:11

At least as much uh common denominators standardization as possible.

Henrik Göthberg: 1:39:16

And then I then I then I think the misconception is that when you when you come to data, you cannot standardize because it's all custom. But that's the what happened now is that what you are extremely standardized, but on much lower technical levels in terms of patterns, yeah, and then you are very standardized in your innovation to adoption approach. So you build the instead of having the governance built into the ERP process, the governance is built into the product development lifecycle or the innovation adoption lifecycle. The standards are built into your hardcore data sharing protocols and data governance stuff.

Helena Hörnebrant: 1:39:58

So it becomes modular pieces. More than process pieces.

Henrik Göthberg: 1:40:02

It becomes modular pieces, that is then where we come back to the point of discipline. That if if you don't do that properly, it doesn't work. But if you do it properly, it scales.

Helena Hörnebrant: 1:40:14

And and here you need like a wonder kid on a block. And I certainly have to do that.

Henrik Göthberg: 1:40:20

No, but one of those. The problem is how do we get people to keep the discipline across uh because I think we are we are cheating ourselves, we are shooting ourselves in the foot over and over again when it comes to standardization and scaling platforms. We're not scaling platforms, by the way, we're building point solutions in the same vendor. Yeah.

Helena Hörnebrant: 1:40:40

Yeah, yeah, yeah, absolutely.

Henrik Göthberg: 1:40:41

So slow down.

Helena Hörnebrant: 1:40:43

Slow down.

Henrik Göthberg: 1:40:44

You know, patterns is one thing where we are extremely standardized. You I can put them patterns in AWS or Azure or Snowflake, whatever. And then oh, we have we all have Snowflake, so now it's great. Yeah, but you all used it differently. Yeah. So you have point solutions using the same vendor. Yeah. What have you achieved? Very little.

Helena Hörnebrant: 1:41:03

Because we I think we need to have reusability. That's the the key word. That's the key word here. Everything we can reuse is good. That was more than two minutes, Mr.

Henrik Göthberg: 1:41:13

But that's now we took an example and we took this the key topics of product perspective, organization perspective, investment perspective, ways of working perspective, platform perspective. Not bad. Not bad.

Helena Hörnebrant: 1:41:27

And we sold the world, we sold the world, you know.

Henrik Göthberg: 1:41:29

So we assault data innovation summit that we okay. Let's go to future outlook. And I framed it when we were discussing what will be different in how we think and work with scaling the next 10 years versus how we work with it the last 10 years. And I I I added something because in between the lines now, the whole conversation as we're following that with the book and all that. The hint is this are we saying that the core objective function, what we should work towards, has fundamentally shifted? We are coming from scaling, understood from efficiency in economies of scale as the number one objective. And now we have flipped to the main objective of scaling is adaptability and flexibility. Yeah. Because this is sort of the entry point to what's scaling the next 10 years. Do you agree with that?

Helena Hörnebrant: 1:42:25

I agree fundamentally. So if you're stuck on efficiency, you will never be able to take any of the emergent technologies in in its uh appropriate way. You will only get very slim results with that.

Henrik Göthberg: 1:42:40

Okay. So if that if we are we are baseline that, so then how is this different and play out differently moving forward now? And this is what I like now, because you actually you've been a very diligent guest that you actually gave me some ideas that we could talk about. So this are so I'm just giving the list. This is not this is Helena's brilliance, not mine. And you did, and I really want to go through them because you did the future outlook exploration, and then you gave me an email back. I'm not sure if it's you or Chat GPT, but that's that's I'm sure it was you. But this I'm just gonna read the list out, and I think we should uh pick them apart a little bit. No, no, no. But you you you gave me from efficiency to adaptability and optionality as as a key heading, I mean as a keyword. And then you said from process standardization to AI native workflows, and then you said from organizational scaling to ecosystem scaling, then you said from linear talent growth to exponential productivity per person is so many good topics here, from centralized decision making to distributed autonomy, from long planning cycles to always on experimentation, from standard change management to human-centric transformation, from internal data to shared and synthetic and real-time data.

Helena Hörnebrant: 1:44:02

We haven't talked about that.

Henrik Göthberg: 1:44:03

No, we haven't talked about so we have touched a couple of these now. What I liked with this is sometimes when you do this before-after, uh like the Agile Manifesto was set up the same way. It gives a feeling to what we're talking about. We are moving from here to here in the way in the way of thinking. We're not gonna pick apart all of them, so let's pick up favorites. I think there's a couple here that I really want to talk about. I think we need to take a little bit more time on the from linear talent growth to exponential productivity per person. Yeah, because this is something that this that's an interesting one, and it and it and it lives on the individual us as individuals and as teams and as enterprises. Uh so I'm just picking one of them. Uh, so this is one of the angles of how to think about scaling that you put down here that are, you know, we have so much content here if you want it.

Helena Hörnebrant: 1:44:56

No, but let's stick to that one.

Henrik Göthberg: 1:44:57

Let's start there. Let's maybe do five minutes and then we'll see how many we can do.

Helena Hörnebrant: 1:45:02

I think that uh we have been way too narrow-minded in how we think about um talent and talent development. I think with the new type of um emergent technologies that we have, we don't talk about. I mean, I literally took on co-pilot for two weeks and I could improve myself 20%, add 20% of the capacity, take away, if you will. But I also realized if I do a bit more, I I took away another 20. But um I think if we use it right and we use agents and build them specifically for my own need, maybe I can become 100%, 200% of myself. That means something so exciting that we have time left to connect as people. It's almost like primary, secondary, and tertiary um um research that we have the basis laid out in front of us. Now it's our turn to connect and see how we can use those insights to do something much bigger. And I think also with um hybrid workplace people need to come in and connect. They don't need to come in and read through a piece of paper and sit on their own corner, they need to discuss, connect, and grow. So I think it's two pieces there. It's for us to get back to human beings to connect and well-being, but it also with the AI and uh all the tools that we have there, we don't have to look at 100% capacity or 40 hours a week. We can have 20 hours a week in capacity for 300%.

Henrik Göthberg: 1:47:01

Yeah. This this lines up really well with the philosophy of first horizon and second horizon AI that Matthias Fross was onto. And even Scorney is doing this as well, where we're looking at horizon one is about the personal productivity frontier. Yeah, you're not really doing enterprise grade workflows yet. No, but this becomes really important because what you said now, and I I even even before the pod, we were we were talking about this privately, me and Goran, around how we are extending ourselves as CEOs for different companies and how we are basically now building our agents around us. Has nothing to do with our core workflows, it's more about how we comportmentalize our CEO jobs in different ways in order to basically build our, you know, what is it you call it, Goran? G2. He's G1 and he's creating G2.

Helena Hörnebrant: 1:47:57

Yeah. Goran, we want actually G2, G3, G4.

Henrik Göthberg: 1:48:03

No, no, no, no, no, no. That's another level. They are called Henry Harvey. What was your name? No, I have different, right? It's GPTs, yeah. And so this is then what we're talking about. That is this is not about replacing a little bit about um work here and work there, it's about extending yourself and it's about building yourself as a team. Yeah, but you are taking in some ways personal responsibility for it. This would then be something else in Horizon 2 when we're going to enterprise create workflows, yeah. But but this is the first step. Yeah. And the way I think what you said now, one of the reasons what why we need to do it is to have more slack to think and more slack to connect with people. Yes. So it's it's so it's like, as you said, we are so slim now. So when we are slim, we don't have time to think and to innovate. No, so this is about creating the slack for us as humans to move up in the abstraction level in order to work with innovation, reflection, thinking. Yeah, so we need thinking doers was Matthias' favorite expression. And how do I create a thinking doer of myself when I don't have any slack? This is what we're talking about.

Helena Hörnebrant: 1:49:14

Yeah, so how do you create room for the next wave?

Henrik Göthberg: 1:49:18

Exactly.

Helena Hörnebrant: 1:49:19

But I think it's not just the next wave. I think this is here to stay. I think this is-we will have expectations on our employee that's far exceeding what they do today.

Henrik Göthberg: 1:49:30

Yeah, and this is the problem, right? Because if everybody doesn't understand that this is the same logic as when you said, Oh, you kind of need to know how to use the PC.

Helena Hörnebrant: 1:49:38

Yeah.

Henrik Göthberg: 1:49:40

I mean, like if you think about it, using a computer to not using a computer, that sort of 10 times our productivity. It's the same thing again. Yeah, it's not different.

Helena Hörnebrant: 1:49:48

But when we started university, you didn't even need a computer the first year. You borrowed at the university and all of that. That's so funny.

Henrik Göthberg: 1:49:56

Computer we were sitting there on Mac, remember that? Okay, so that's a good one. Let's keep that one. Which one? I mean, like organizational scaling to ecosystem scaling. We kind of touched upon that.

Helena Hörnebrant: 1:50:09

Yeah, we have touched upon that.

Henrik Göthberg: 1:50:10

Okay. Um, we haven't said so much from internal data to shared data. We haven't said so much about that.

Helena Hörnebrant: 1:50:17

Let's talk about that because I think that's a super interesting one. And and to be fair, uh, there's a lot of there's probably a whole program we should use for this. So let's do a cliffhanger and see where you invite another smart person to discuss this. This is a truly, truly, truly interesting topic. Do you want to start? I think we all have ideas on how we approach this. So we can basically say that in the in the past we scaled rely on BA dashboard, historical data, uh, we had siloed data, we abstracted them in different ways. And and now, of course, we talked about data mesh. We don't have to talk about that right now here, but we also can use synthetic data.

Henrik Göthberg: 1:51:01

We can use other ways, yeah.

Helena Hörnebrant: 1:51:02

Industrial data that we buy in, we can do digital twins of ourselves or other people, we can do or uh power scaling in a completely different way.

Henrik Göthberg: 1:51:16

Okay, so let me my angle here. I mean, like we even had that, this was the keynote. This was the day one or day two keynote of data innovation summit, where we were referring and talking about this as the intelligence plane. You know, so so basically we are we are saying this is going whore for anyone interesting want to go nerdy, this goes all the way back to cybernetics thinking and the feedback loops. And and the fundamental topic of how to develop intelligent, or intelligence is hard to define, but we can always talk about intelligent behavior in an ecosystem or in an enterprise, is that whatever you're doing is functioning effectively in relation to everybody else. So you get to a point where you have a process plane, which is the transactional flow, the linear flow of goods or whatever. Uh, and we're gonna now have a credit contract, blah blah blah, you know. So that's the linear flow. When it comes to people and humans and how teams work and how we really make decisions from very uh operational decisions to very strategic one, it doesn't live in the process flow, it lives in the intelligence flow, it's feedback. And in a normal organization, those feedbacks is anything from email to phone to something in data like this, but it's it's it's a human uh it's it's a human game. All that needs to be encoded, all that needs to be data and encoded data, so you have a completely different ball game of how these things that is the normal interaction needs to be encoded. Yeah, why? Because one of the coworkers in the team or the whole team is AI, yeah. And all of a sudden then the AI doing their job needs exactly the same feedback loops as you as me as people need to make decisions. Yeah, so this is huge, right? So that that it means nothing is internal data anymore. What do you mean, internal data? Yeah, it does it it's it's not that simple. Internal data in one silo, internal data is all of so you need to have perfect signal flow of every data piece in Scania. That's insane.

Helena Hörnebrant: 1:53:22

I I think that's insane. So break that down so people understand what that means.

Henrik Göthberg: 1:53:27

No, because it's a little bit philosophically, I would argue like this. I would if I should build a perfect brain, intelligent brain of Scarnia, do I envision that as some sort of monolithic central brain in the middle and then use everything else is dumb? Or do I envision it as a hive mind? So each and every cell is intelligent, and the what makes the greater more intelligent than the part is a perfect transparency and signal flow. So the way data flows from signal flow from the lowest cell to the to the brain and back makes us intelligent. Yeah. And what I mean with that is that the the human body, really. The human body, right? Because if you think about it, if you if you if you if you look at the enterprise as an organism, the latency of sending of making decisions on the top for every little micro decision on the bottom won't fly. So everybody needs to autonomously be able to, within their context, make decisions. Those decisions that they made has consequences for for the rest of the body. Yeah. So all of a sudden now we need to have we need to have a signaling system where I made this decision now. Oh, we have we we are we're having a shortage of of uh of semiconductors. So now I made this decision to and that has huge the whole chain in order to deliver has impact on that. Yeah. So orchestration. So orchestration then becomes fundamentally signals data that needs to be at real time shaping each other. Yeah, so this is cybernetics, right? And this is we know what we then call that the intelligence plane that we we can now understand all the work we're doing as decision data workflows as to put a label on it, very operational or very strategic. We have yeah, what is the decision data workflow around the strategic investment planning of Kanya? Is one here, which is many different cells underneath, and then you can have what is the decision data workflow around the incidents when we have a uh shortage in one of our suppliers, yeah. All this is different signals that has impact around each other. So when I so from internal internal data to shared, real-time data, even synthetic or whatever things, in order to understand that we make as good decisions in relation to whatever this can be very, very granular. It's it's a fractal problem, right? Yeah, it is um that's to me that's what it means. I mean, like so. Now I've spun my angle onto it. I don't know, you can take it back.

Helena Hörnebrant: 1:56:10

No, I think it's fine. I think we we've spent a lot of time about it, but it it's really about forecasting what's happening in every cell. Yes, or in communicating, yeah.

Henrik Göthberg: 1:56:22

So I it's both ways, by the way.

Helena Hörnebrant: 1:56:24

Both ways. Yeah, both ways. So the signal system and the simulation goes all the way in and back. And and I think what we did before is just to base everything on history. Now we need to base it on everything that we know going forward as well. But this and we simulate.

Henrik Göthberg: 1:56:41

But this becomes a nervous system, and I argue that the nervous system is not the report in the old way, it's the human, the real-time human connection. The report is helping the human nervous system to make these decisions and to understand the world. And it's it's a limited data, it's fragmented data, and it's reporting data that is stale and old. Yeah. And we are we are now needing to figure out that nervous system encoded in real data.

Helena Hörnebrant: 1:57:08

Yeah, it is. And I think uh um in any case, it doesn't matter if it's our internal data or not, data is everybody's business. Yeah. And how we set up the analysis models is everybody's business. So it becomes something uh very um well, it should be on everybody's uh top agenda, no matter where you work.

Henrik Göthberg: 1:57:32

Yeah, but and and we now then need to understand who is working on it and how are they working on it on the platform level. And then, of course, when we get to the real data decisions, the real decision data work that sits right in the workflow.

Helena Hörnebrant: 1:57:46

Yeah, I think so. And and but I think if we if we summarize what you just uh said with everything that you said, it is that connecting ourselves gives us leverage. Yes. And even more so, uh getting the external data in becomes a true competitive differentiator for anyone. Yeah, it's how we combine it and make insights on it. So from internal data to shared, uh synthetic and real-time, both external uh and synthetic data.

Henrik Göthberg: 1:58:19

And by the way, that will be the fuel for any AI we want to bring in.

Helena Hörnebrant: 1:58:24

Yeah, that would be the fuel for any AI we will bring in. Okay. It's so true.

Henrik Göthberg: 1:58:28

But then let's take, we are always up onto two hours. So I wanted to take a couple of macro questions. And if we now talk about, you know, we didn't we we've talked a lot about the scaling dimension. So I asked the listener to a little bit think about all the dimensions we are talking about, credibility, trust, you know, all this and and and those things. Can we say something about these this new type of scaling and these dimensions we we used to discuss? What do they mean for Sweden and Europe in the age of AI? You know, what do we need to do differently if we want to scale up and be successful in the age of AI? And what I mean with scaling up, it can be many different things. It can be to scale out the you know, on the productivity frontier, industrializing the use of data and AI. But but more importantly, is to have thriving scaling businesses at large at this age, when when everybody needs to then grow or deal with this. Yeah, even the public sector needs to, in a way, yeah. I mean, like we have a shortage of of uh people going in retirement soon. So we need to scale up, we need to be able to do produce less with more there too, right? Yeah, yeah. So so any thoughts on this more macro political macro view on on what becomes important or what we should think about.

Helena Hörnebrant: 1:59:48

I mean, of course, everybody who hears the word Europe hears regulation. Yeah, we're gonna continue to regulate, it's gonna be heavier, but hopefully we can ask. Can request them to do it smarter. Not just, you know, black and white forbidding, but rather tell us how we can apply it in a smarter way. Um, I also welcome, if it's possible, to have labs somewhere, somewhere where we can actually try out things without getting punished for it. So certain sectors that we need to move faster. It can be diagnostic, for example. It could be uh schools obviously have to adapt to the AI world. What is learning gonna look like in the future and what would that be, etc. Uh, but I also hope that we can put at scale the human more intelligent version of this uh to get help in our workforce because we are way too many old people in comparison to uh workers. So I don't think um we can't handle that on ourselves. Uh and if we look at China and if we look at the US, they're way before us on this one. So see if we can at least find labs in Europe so you can go crazy a bit on this one.

Henrik Göthberg: 2:01:11

But but let me test one angle that I think ties into this, and actually I'm I'm triangulating what we are talking about here with what we talked about with Nordia uh last week. Because many of these things, when we want to move fast or more, move move smoother, comes down to from a regulation point of view, legal uncertainty. Yeah. And I could I could argue that even why we move slower or faster inside our enterprises, we could we we we prove that within Nodea, that yeah, a lot of times if you want to do proper innovation with proper governance, so so as Matthias said it, if I want if I don't if I want to do innovation without governance, I should work in a startup. If I if if I want to do innovation in a bank of Nodea size, I need to embrace governance. Yeah, what it means with that is a systemic design, yeah, and the and the process we need to trust the process. And be and then he says legal uncertainty on regulatory is one thing. We don't know what good looks like or how we should do it. To some degree, we have the same problem with seeing the enterprise. We don't know what good looks like, or different functions argue, or the security function says you should do it this way, yeah, the other guy is that way. So all this boils down to what I think is one of some of your core principles. We need to be way better at systemic design, we need to be way better at continuous redesign. And that those kind of principles, your keywords, how do you get them into EU? How do you get them into you know that's equally important? Yeah, it is, and even in Sweden, right? Yeah, it is because we are now talking about the E AI strategy of Sweden. We need to do this and this and this and that. And I hear people talking about investing the money in the infrastructure and we need to build this and this. And I've been trying to get my foot in the door on a couple of things because I want to talk about your principles, I want to talk about your way of looking at steering because I think it's a blind spot in the whole conversation. Yeah, so we are we are not fixing the fundamental design of the arena, we are used trying to push in more money or more tech, yeah, but we haven't fundamentally redrawn those principal ideas. Yeah, what do you think about that rant? That's a rant, of course. But am I so that's how that's how my connecting back to what you said? What we need to be having labs and we need to have regulation, we're gonna deregulation. But what is the problem is that we don't know how to how to navigate. If I knew how to navigate it, it wouldn't be a problem.

Helena Hörnebrant: 2:03:39

No, and I think we need to navigate this with um advisors to the people who put the regulatory frameworks in place.

Henrik Göthberg: 2:03:48

They need to know more about how this really works, right?

Helena Hörnebrant: 2:03:50

They need to know more about how this really works, and I think we need to, as corporates, also lend out people to that and see how we can solve certain things. Co-create. Yeah, co-create. And I think we need to really look at the uh societal uh values and see if we can solve them together across Europe as well. Uh, but but I I think we need to be also good at working closer. If we can't work closer to China, we could actually work closer to India. That's speeding up this quite a lot.

Henrik Göthberg: 2:04:24

I I have I have one core topic, which is almost like a cliffhanger for another pod. I mean, like there's a couple of pods or discussions that I've had with you over the years that we have had them in the pod here as well. But with you particularly, we've had this conversation. And I mean, like we we can see in Sweden and Europe that we have a strong startup scene in all this. Yeah. But I would argue we we don't we are not so good at hyperscaling. And even the hyperscalers that we have seen come out of Sweden, like like uh Spotify and all them, eventually they're more American. I mean, like it's the uh American venture capital that sort of takes over and then we go that way. Or even when we do an IPO, sometimes we decide to do the IPO in US if we're big enough. And if you now go more closely into this and looking at uh the VC or the PE industry and large enterprise, how we think because I think the Swedish and to some degree European legacy is scaling through large industrials like Scarnia, like Ericsson, like Bosch, like uh Mercedes, right? That's the backbone of European the the European industrial complex or something like that. So just to make it interesting, everybody complains that oh we don't have the right capital in place. And okay, so let's take that out of the equation because that's a whole topic. That's a whole topic in itself, and it and it and and I think it's quite obvious that right now we we we don't know how we are not able to manage to muster the capital or compete. I mean, like when you can get 10 times amount for your company in in US, why wouldn't you? Right. But if I take that that thing in its own, I just take that off the table. What other mechanisms or what other things should we work on in Europe? Uh, you know, if if if I'm a PE person or a VC person or if I'm working in enterprise, you can probably not hyperscale in Sweden.

Helena Hörnebrant: 2:06:36

We have too few people. Yeah, so you need to, if you're gonna scale, you you still need to use other markets. But I think we need to find ways to scale smarter, yeah. Uh and uh depending on what that is and what kind of product you have, it's more personalized, or it can be used in many fields, like um a lot of uh the old Nasdaq technology went straight into the um technology of uh gambling.

Henrik Göthberg: 2:07:06

Interesting, right?

Helena Hörnebrant: 2:07:07

Yeah, and that technology later uh came into um uh uh gaming. So, how do you move your technologies into different areas? So I I I I I don't have answers on how to scale because that depends on the actual profitability model.

Henrik Göthberg: 2:07:28

But it's something we talked about in the past, and I think it we actually talked about it on this pod as well. The dynamics of of scaling in Europe is of course not one central market. No, so if you look at the the place, the game, the that that sort of China takes the approach, we are implementing AI, and they have sort of a a way to use code that way, and America still has an enough one market so they can sort of scale up an innovation and and then you know forget about the regulation, use go hard and break things, as the meta as Ackerman said. In Europe, we never we we our playing field is distributed, yeah. And I always I I get it, I I I almost get the I get the um I'm thinking about how you are talking about this. Uh how can we do central muscle in terms of enterprise muscle in in Trottel Financial Services, and at the same time be very, very flexible and making something work for a very small local market. So that whole bimodal idea, that whole thing that is in the DNA of how what you need to think about on a daily basis, I think that's some that is somewhere here, is like put a European model. Yeah, that is something else.

Helena Hörnebrant: 2:08:55

I don't know if there yeah, it's absolutely something else. Let's uh just make that clear. Um, I think uh uh if you take it from another perspective, customers, we as human beings have almost the same problems to be solved. Yeah, so as consumer goods, I think it's easier to approach us as humans. Okay, but when it comes to making business in certain fields, it's more local than it's European.

Henrik Göthberg: 2:09:21

So you you think low would say global, think global, act local. Yeah. And and that is that is you you cannot get away from that in Europe, I think.

Helena Hörnebrant: 2:09:33

No, I don't think you can, but if we take Spotify, for example, go straight to consumers. Okay, that that they have some regulations. But go straight to consumers. Uh so I think the hyperscalers are more consumer-based, unless you can take a consumer product like payment and make it into companies as well.

Henrik Göthberg: 2:09:53

So so it also becomes an interesting topic hyperscaling in the terms of B2B to B2C. So Mike Lorna then. What are they B2B or B2C company? They're both.

Helena Hörnebrant: 2:10:03

They're both right. Yeah, and payment flows is a bit different from um uh because we have a global banking system. We have a global banking system, but I think they have gamed it uh so they take care of it all behind the scenes. So it's really how they connect their ecosystem and how they grow through their ecosystem apropos scaling in ecosystems.

Henrik Göthberg: 2:10:26

But then you have also some sort of you can learn some a lot from looking at the Klarna model in terms of how they took the ecosystem and how this they created something way simpler as a way to go to the local market or the local small buy, small enterprise shop and still make it work.

Helena Hörnebrant: 2:10:44

Yeah. And I think that's how we learn. You have to go, uh we can't look in our own industry only. We need to look cross function of cross-industry to learn more.

Henrik Göthberg: 2:10:55

Right. Interesting topic. Let's let's see if we should have a panel on that, really. We should have a panel on that. You should need a panel on that because you need a couple of different perspectives when you discuss that question. We do.

Helena Hörnebrant: 2:11:06

So we should invite P people in and uh go around.

Henrik Göthberg: 2:11:10

I have a panel. We have a we we want to have a panel on on these topics. We need to figure that out for next year. I have a panel. So you have a panel, and you need to invite her to that panel. Yeah. I think we should wrap up. Yep. Um you remember the last time when we had this, we also ended up with a panel. That's good. I like when we do pods that helps us book the next pods. It's beautiful. Okay, so that that's a good question, actually. We we always forget to ask the question. We we have we have I have two wrap-up questions now. I'm gonna go straight to the to the crunch. We have a very standardized question that that almost is like a uh a poll or like a research. We all we have asked the same question now for at least 50 to 100 guests. And we want to we we just want to see how people think about that. It's just so it's it's one of it's the port's final standard question. So I'll take that first and then I'll uh then I will do a bonus final question after that. So two questions left. So so the the final question is all the way out into the AGI space. You know, what however you want to define AGI, do you think we're going in that direction? Do you think this is science fiction, or do you think eventually we will get to some sort of uh view where we have AI out there that is equivalent to any human worker doing any human working task? At least, I mean, like you can think about this as uh office work in the digital space and then even beyond in the physical space, yeah, like robots. Do you do you believe that we will see that at all? Or is you know, so no time.

Helena Hörnebrant: 2:12:52

You and not too far away from the future, but I don't think in every single space.

Henrik Göthberg: 2:12:58

No, but but we definitely further okay. So if if we put that AGI topic here, and then then I draw out a spectrum, you know. So how would that world look like in five, ten, twenty, thirty, fifty years? And the spectrum is of course, then you have the AI doomers and dystopian view here that every goes everything goes to shit, and we have the Terminator and Skynet world, or all the way to utopia, a life of abundance, you know, those pictures. So we have Nick Boostrum, we have Max Tieg, Mark, we have many different thinkers, philosophers who've been talking about.

Helena Hörnebrant: 2:13:34

I think we're somewhere in between. We're gonna have some really destructive side of this, and we will have some really fantastic sides of it.

Henrik Göthberg: 2:13:41

Yeah, and oh, so you you are saying both.

Helena Hörnebrant: 2:13:43

Yeah, of course.

Henrik Göthberg: 2:13:45

I like that because my my core question is a little bit like where on the spectrum do you put yourself? So that was that was my step one.

Helena Hörnebrant: 2:13:53

Yeah, so I ended up saying both.

Henrik Göthberg: 2:13:55

Yeah, you together with uh Sverke Jansson, head of AI research at RICE and a couple of others, I share that view as well. We have always through society have the riches and the poor, and we had had the wars over here, and we've had the yeah, like so to think about a world in perfect abundance or perfect harmony, or perfect, you know, it doesn't work like that. So both, right? So then you then my core question is how can we make sure that we are gravitating towards the good, or how can we how can we never think, you know, what is your recipe to think to to be a contributor for good, AR for good Henrik?

Helena Hörnebrant: 2:14:39

If I could solve that question, we would solve the world's biggest and finest questions. I have no clue, but I I I hope we do more of the good things. Yeah, I hope we fill up uh our world and people's worlds with uh more of the good.

Henrik Göthberg: 2:14:57

Yeah, my my simple answer to that is that it's up to us to shape what good looks like. It has to it's it's up to us to take the good fight in every micro topic we do. Yeah, from the smallest thing we build, that we build it right ethically, to you know, thinking about uh pushing the lawmakers in the right way, stuff like this. But in every single step, anyway. Yeah, love that answer, double. We it's both. Final question then. If you were to invite someone to the pod, or someone you would like to see uh in the pod on on a topic, or someone you think we should have on the pod, do you have any recommendations for us?

Helena Hörnebrant: 2:15:40

I have many recommendations, but does it have to be a Swedish person or should be a Swedish person?

Henrik Göthberg: 2:15:46

It should be someone who can travel. We prefer to do it physically, so that's the only limitation.

Helena Hörnebrant: 2:15:52

Yeah.

Henrik Göthberg: 2:15:52

Uh we don't we have no problem having someone coming joining us here, but it's more about uh what is feasible.

Helena Hörnebrant: 2:15:59

I don't know if I have a specific name because there's quite many, but I would really like to see AI in creation of uh art.

Henrik Göthberg: 2:16:07

AI and art. We've had that.

Helena Hörnebrant: 2:16:10

There's a Finnish um carpet maker that's done fantastic carpets with AI, but I also think that music and to see how we incorporate that. I mean, what is it? 30% of Spotify's music is made by AI. I don't know.

Henrik Göthberg: 2:16:27

Yeah, we've had a couple of those angles.

Helena Hörnebrant: 2:16:32

I'm glad you're saying it. I'm stretching it and exaggerating. So thank you.

Henrik Göthberg: 2:16:36

No, but AI and arts, AI and music, stuff like that would be interesting, yeah.

Helena Hörnebrant: 2:16:40

Would be interesting. The good bits about uh AI.

Henrik Göthberg: 2:16:43

Yeah, yeah, for sure.

Helena Hörnebrant: 2:16:46

Thank you.

Henrik Göthberg: 2:16:46

Thank you so much for this, and uh we we will maybe talk a little bit after after work. Take care, thank you so much. Thank you.