Rise of AI Conference 2025 · Berlin

AI for Society: Shaping a Positive Future

Elisabeth L’Orange, Kenza Ait Si Abbou, Annika von Mutius, Nikolaus Roettger

At Rise of AI 2025 in Berlin, a panel with Elisabeth L'Orange (Deloitte), Kenza Ait Si Abbou (FIEGE), Dr. Annika von Mutius (Empion, KI Bundesverband) and moderator Nikolaus Roettger (FutureStory) discusses how AI can benefit society beyond the circle of people already working with it. The panelists trace their own paths into AI, from academic mathematics and a university neural-network class to consulting and company-building, then debate whether the wider public needs to be excited about AI or simply needs enough trust and literacy to use it well. Von Mutius argues most business users do not need deep technical competence, only clear economic value, citing Germany's roughly six to seven percent AI adoption rate against far higher figures in the US and China. Ait Si Abbou stresses that legacy IT systems, not lack of enthusiasm, are the real barrier for many German companies, and that involving end users in building tools drives adoption. The panel also discusses Klarna's reversal on replacing call-center agents with AI, the risk of AI-amplified disinformation, and closes with differing personal optimism scores about AI's impact on society.

Key takeaways

  • Kenza Ait Si Abbou says legacy IT, like FIEGE's 151-year-old AS400 warehouse system, blocks AI value more than lack of enthusiasm.
  • Annika von Mutius cites a roughly six to seven percent AI adoption rate in Germany, versus about 60 percent in the US and China.
  • Ait Si Abbou notes cancer diagnosis accuracy rises from about 60 percent for new doctors to 99.26 percent for a recent AI model.
  • Ait Si Abbou says Klarna's reversal on replacing 700 call-center agents with AI shows adoption will 'zigzag' rather than progress in a straight line.
  • One panelist says the disinformation network Pravda got ChatGPT to produce false claims about Russia in about 30 percent of cases reviewed.

Detailed notes

How the panelists arrived at AI

Von Mutius started working with AI systems around 2016 while finishing a mathematics master's and PhD, then worked in Silicon Valley before founding Empion in Berlin, a recruiting system that quantifies values, company culture and personality traits to match people to companies; she says she chose Europe over the US because of its value system, despite enjoying the enthusiasm for venture capital in San Francisco. Ait Si Abbou traces her interest to a university class around 2000 where a professor used cancer diagnosis with neural networks as an example; she wrote her 2003 thesis on clustering algorithms and says she was told at the time that AI meant an academic career, not industry. As CTO and board member at FIEGE she is responsible for IT, data, AI and information security, and describes her main work as reducing system complexity and building the infrastructure that AI needs to scale. L'Orange studied law, wrote a master's thesis on algorithmic trading using Goldman Sachs's algorithms, then founded a company, worked as a VC investing in deep tech, was a CFO, founded another company, and now leads end-to-end AI transformations at Deloitte.

Does the public need to be excited about AI?

Von Mutius argues that excitement is not a precondition for using AI, since many business users do not even realize AI is inside the tools they use, particularly for predictive systems; what matters, she says, is that a tool has demonstrable economic or societal value, for instance better time-to-hire and retention in recruiting. Ait Si Abbou counters with medical diagnosis figures: doctors leaving medical school diagnose cancer correctly about 60 percent of the time, experienced chief doctors about 87 percent, and a recent cancer model reached 99.26 percent accuracy, she says. She warns of a widening gap between AI-fluent people and those left behind, and cites a figure that 85 percent of pilot-to-project transitions in companies fail because organizations and employees do not adapt. Von Mutius adds that Germany's AI adoption rate is around six to seven percent, well below roughly 60 percent in the US and China.

Legacy IT as the real bottleneck

Ait Si Abbou describes FIEGE, a 151-year-old company, as running its warehouse management on an AS400 system programmed in RPG, so startups' plug-and-play AI pitches assume an API that does not exist; building that infrastructure has to happen before any value creation starts. L'Orange says the pattern is widespread in Germany, recalling a fund she worked for that still used 22-year-old CRM software and colleagues on Word 97 only a few years ago. Both frame the legacy debt as an opportunity: healthcare systems in parts of southern Germany that never digitized paper-based processes could skip straight to AI rather than going through a horizontal digitization phase first.

Building the bridge between the AI bubble and the public

Asked what connects the AI bubble to the general public, Ait Si Abbou points to writing accessible books, starting in 2020 while at Telecom, after running internal 'AI for everyone' workshops where employees arrived afraid and left relieved once she explained AI as applied mathematics such as regression. She recommends the free 'Elements of AI' online course developed by the University of Helsinki and translated into most European languages. Von Mutius adds that showing AI's weaknesses, referencing an AI incident database of failures, and comparing generative AI's output to a philosopher like Sartre, can reduce public fear of AGI or superintelligence scenarios she considers still far off. Ait Si Abbou also stresses involving end users, such as call-center staff, in designing tools from the start so adoption spreads through peer advocacy.

The Klarna reversal and non-linear adoption

Ait Si Abbou cites Klarna's announcement, around six months earlier, that it would replace 700 call-center agents with AI agents as part of its IPO story, and its recent partial reversal to rehire staff, arguing roughly 20 percent of call-center cases will always need humans and that running proprietary voice AI is too costly for many companies, which instead pay per-call fees to providers such as ElevenLabs, pushing large firms like Lufthansa and Deutsche Bank to build their own synthetic voices. She concludes that AI adoption will 'zigzag' rather than progress in a straight line, and that predictions of imminent AGI or superintelligence from tech executives are overstated.

Closing outlook on AI and society

L'Orange highlights AI's potential in healthcare and in translation tools that let people follow television or school curricula in their own language, including for migrants and children. Von Mutius is most optimistic about reallocating the German labor market, where many people sit in the wrong jobs despite high employment, and about AI's role in defense technology using real-time data, calling the topic more than critical for Europe currently. Ait Si Abbou expects AI to help address climate change, migration and resource scarcity by finding new materials and efficiencies. Asked to rate their optimism about AI and society from zero to ten, one panelist answers ten, L'Orange declines to give a single number, saying AI is a tool whose outcome depends on who uses it, and another panelist says five, citing a case where the Russian disinformation network Pravda got ChatGPT to generate false claims about Russia in about 30 percent of cases reviewed.

Quotes

It's not going to be a linear situation. It's going to be like zigzagging.

Kenza Ait Si Abbou

AI is not something that is just there because it's AI, it must have an economic value.

Annika von Mutius

We shouldn't forget that we are in the driver's seat.

Elisabeth L'Orange

Questions answered in this talk

What is blocking AI adoption in German companies?

According to Kenza Ait Si Abbou at Rise of AI 2025, legacy IT infrastructure, such as her company's 151-year-old AS400 warehouse management system running RPG, is the real barrier, not a lack of plug-and-play AI tools.

How accurate is AI at diagnosing cancer compared to doctors?

According to Kenza Ait Si Abbou at Rise of AI 2025, new doctors diagnose cancer correctly about 60 percent of the time, experienced chief doctors about 87 percent, and a recent AI cancer model reached 99.26 percent accuracy.

Why did Klarna reverse its decision to replace call-center staff with AI?

According to Kenza Ait Si Abbou at Rise of AI 2025, Klarna announced replacing 700 call-center agents with AI ahead of its IPO but later said it needed to rehire staff, since about 20 percent of cases still require humans and running proprietary voice AI is costly.

Mentioned

Elements of AI, AS400, RPG programming language, generative AI, algorithmic trading, AI incident database

Keep exploring

More from Elisabeth L’Orange

More from Rise of AI Conference 2025

More on Enterprise AI & Adoption

Transcript

Auto-generated transcript, not edited. · 7,829 words

Show the full transcript

So we are now in the final stage of um the rise of AI um here on the meta stage til what you're saying. So we will have uh one more panel and this is moderated by Nicolas Ratka he's the founder and partner of future story advertising CEO CMOs and leaderships teams in strategy and please welcome him with a round of applause. Thank you very much my friend. Thanks. Yeah, it's great. It's great to have you here and we continue the discussion we just had about good news from Germany. Now we want to talk about AI and society and how it can have a positive impact. Please give a warm hand for cancer a zu Elizabeth and Dr.

Anahio will come. How great to have you on stage. Hi cancer Elizabeth. Hi, welcome. Hello, Anukica. Nice to have you here. So, it's great to have you here. Yeah, let me introduce them to you. So, next to me is Elizabeth. Hi, great to have you here. You're a partner at Deoid. You're focusing on AI and data. And before joining consulting, you had your own AI company. So you have the consulting, the corporate and the entrepreneurial view on our topic today. Welcome Elizabeth. Thank you. Next to we have Kenza. It's great to to have you here. You currently work as CTO and boardman member at Figer. Um we already heard in the last panel that you're an expert formemes and kamoose and can tell us a little bit about this.

Um you have written several books on AI. The last one I think was an emotional AI. So we will figure out what this is. It's great to have you here. Welcome Kenser. Thank you. And we have Dr. Anika Famutz with us. Hello Anika. It's great to see you again. We met at several conferences during the last last two weeks. Um you are a board member at Kibbundes at the German AI association and co-founder and co-CEO of Ampion and HR tech startup. What you exactly do I will ask you in a minute um because um I thought when you have been at the last panel about good news from Germany we discussed a lot about um companies and corporates.

So I thought we get more on this topic on a on a more maybe personal level and talk about a little bit how you all not only started with AI but what you now do with AI and then in the second step how we can use this to persuade others in our society and to talk with others that are not in our AI bubble at the moment uh to get the story out why AI is something good something we can be maybe optimistic about. So Anika, how did you start with AI and what is your company doing? Uh so I originally started working with AI uh systems back in I think 2016 or something when I finished um my masters and started my PhD back then in mathematics and um mathematics is very abstract in a way which I realized very quickly and then I was capable of doing a job in in Silicon Valley um which I did before I came here to Berlin to found my own company and there we were capable

of applying the very meth mathematical models with uh data. So back then in a very sexy way predictive systems and I very much liked that. Um I very much enjoyed that and basically from all those research and all that um modeling we made our research transfer now to our company today that I founded in in Europe in Berlin which was actually the other way around that most people do going from Germany away to the US and I did it the other way around. So, but I'm pretty happy here and we are working. We are as well. Yeah. Yeah, we are as well. Um and um so what we do here now to why just one question why did you choose this way?

Yeah, that's also what people ask me very often. Um so um honestly sometimes I also don't know but um most of the time I um I so I lived in in San Francisco for over four years in total and I loved the enthusiasm and the uh the the excitement for innovation and also for uh high risk capital the venture capital invested into those asset classes. But um I think that we have an outstanding value system here in Europe. I think we have a wonderful culture. We have wonderful people, wonderful smart people. So apparently over time I felt a little bit annoyed about the let's say capitalism that is lived in the US. I I mean don't get me wrong, I I think I'm a capitalist by heart, but I still think there's more to it than just the money.

And it was all about the money. So I kind of like living here in Europe. Yeah. Maybe that's the reason why you work with values. Maybe. Yeah. That was that is what your company is doing, isn't it? Yeah. Yeah. Exactly. That's what we do. So what we do is we um develop a recruiting system that quantifies values, company culture, personality traits and it's all about making um maximum individualized job search available to the mass market. Um, which means we bring the right people to the right companies uh in order to make people as happy as possible because we have such an interesting system or such an interesting labor market here in Germany. We have uh more people in employment over the past years but less economic output which means that people are working less or they are not working good enough which then infers that the wrong people are at the wrong positions and that's a huge challenge to solve.

Yeah. Okay. Cool. I come to back to this later but I learned you started with AI because of academics and your first job. How did you start? Similar experience actually. So at university I studied electronics and I remember very well still in my first semester uh I had a class on neural networks and our teacher or professor um used an example of cancer diagnosis uh to to make it attractive that with neural networks you can diagnose uh cancer and I was totally oh wow is this possible? I never thought that mathematics could could be so useful and I love maths as well. I'm a math nerd. Uh and that got me really excited. So I'm talking about the year 2000.

Um so I picked all the machine learning, neural networks, fuzzy logic, all those all those um uh subjects. Um and we didn't really use the term AI back then. I wrote my thesis about clusterin algorithms. That was in 2003. Um anyhow, so that was my first touch point back then and that was the interesting part as well. Um after I finished my uh my bachelor degree, I was applying for a master degree and it was either in AI and that would have been totally academia path or uh do project management and then go in in the industry and and I was discussing with my professor back then what should I do and he said okay if you pick AI you should know you're going to academia because companies industry is not using this technology it's still too early um and because of some other private reasons I moved to Berlin and did my masters in in um um industrial not not industrial virtual vision something only that exists in Germany uh business engineering so and

this is where I am now and what is your daily work with AI nowadays well as a CTO obviously I I have very very broad responsibilities So I am um I also have the IT and the the the group IT. So I I need to take care of our infrastructure uh uh it tech data AI all of this um is is um my team and as well as data privacy and information security and I remember in my early days I always said I don't want to be CIO. So I managed to become a CTO. What I didn't think of is that the CIO would report to me which means I'm still responsible for it. Um anyhow so in this area of responsibility I need to make sure that we have the foundations that we're not not just building the shiny use cases or the PC's but obviously the infrastructure and the need needs to be there.

So we're talking really uh the foundations building integration between IT uh systems standardizing processes so you have fewer IT uh uh systems. Our uh tech um or system environment is really really really complex and uh heterogeneous. So I'm trying to reduce that complexity um an integration u um building integration and then the connection to our data infrastructure to give a chance uh for scale in AI. Yeah. Cool. Thank you. So I learned academics plus excitement. Maybe maybe that's a combination to get AI out there. What was your first connection with AI? How did you start? What also academics? Sorry. Alo academics. Yeah. No no news here. Um, no. So, I studied law and then I did a masters as well in New York and I wanted to go into finance and I researched on algorithmic or I researched that algorithmic trading.

I wrote my master's thesis on algorithmic trading and especially I looked into Goldman Sachs algorithms and how they move markets. That's very interesting, you know, because they have a high um have high volume trades and so that's that was my first connection and then I went um founded my own company and then I was a VC and we also uh heavily invested into deep tech. So my focus was deep tech then was a CFO at a company where I also did uh we also did deep tech and now then I founded my own again and now I'm here I'm there at Deoid and we're doing endto-end AI transformations for companies. Okay. So that's your daily work with AI at the moment.

Yeah. So why did I ask this? Because I wanted to figure out how we can excite more people about AI that are not in our bubble. So maybe we have to substract academia because not everyone is going into academia. So excitement is maybe that was left that excited you to start with AI. How do we get the excitement out out of our bubble and to excite more people about the positive impacts of AI? What are your thoughts, Anukica? Uh, I I'm not even sure if everyone must be excited about AI because AI is not, you know, I mean, it can be a hobby, especially when you develop AI, like from our developer perspective. Of course, it may be helpful to be excited about it, but especially if it comes to application um and to users very often um especially for B2B tools, people don't even realize that there's AI inside the tools, especially when it comes to predictive systems.

I think what um happened in the past one or two years is that we realized a hype on Gen AI a lot which is of course only a part of um artificial intelligence and I sometimes feel that people don't need to be specifically excited sometimes they don't even have to have any competences uh to use AI systems and that's something that many people don't know many companies are developing AI concepts in order to implement specific steps after steps and usually Not usually, but in many cases, especially in in B2P applications, that's just not possible. So, of course, excitement is great, but I think it's not necessary in order to apply AI or to just make the best productivity out of what we can get.

Um, yeah. What are your thoughts, Ker? Um, I think the focus maybe is on the value that AI creates at the end. So even in my case, I studied electronics because I I loved mathematics. So I didn't think of AI at that point. But having that example that you can use neural networks to diagnose cancer, this is what got me motivated for the topic. So I I I majored in in um in it. Um and I think for for the broader public when they see the benefits in using something they would just use it regardless of the technology behind. Obviously we don't want people just to use the systems. We want them as well to design and co-create the systems because this technology is there for everybody to use and uh we should democratize it and just not keep it for a small share of the population.

Uh so we do want everybody to learn something about the technology and to be able to to participate in a way not just as consumers uh but I would say as um um wellinformed and responsible consumers and with wellinformed and responsible I mean also giving feedback and and and trying to shape how those products are uh being built and Yeah, especially talking about the morals and values in Europe. What are your thoughts? I I agree. I think the application layer is sort of the interesting thing. If you look into medicine, just like you said, um we see that the current models are have a much higher um um how do you say like they're they're they hit sort of the diagnosis at a much better rate.

So if um a doctor if a doctor comes out of law out of med school um they usually diagnose with a precision of 60%. So they see a cancer case and 60% managed to diagnose it correctly. Then when they progress along along their sort of experience scale 87% of all chief doctors of all our you know super experienced doctors diagnose the cancer right but if you look into the models like into medge ggemini or like even chatbt is is more accurate then it's over 90%. And the current models a few weeks ago a a new cancer model was released they had a 99.26% accuracy. So I think that's amazing, right? Because like if you're if you have cancer, those that little delta, those little 7%, they mean a lot to you.

That could be a life or death sentence. So I think once the general population understands that how AI can kind of save your life, you know, how AI can can make the world more efficient, better, you know, AI will help us save resources, AI will help us teach our kids. I mean, my kids use Genai on a daily base. They've exchanged Google for Chad GBT which is from a from a um resource perspective outrageous because every time you generate a picture it's like five liters of water basically or four depends what you do. So I think like I think at the moment we we see a huge um a huge like delta opening up because we see at the on the wrong on the one side we see the um population that is very AI sort of fluent you know like it's very digitized and it makes us all much better.

It makes us more efficient. You know, we write our emails faster, we we take faster decisions, everything gets better and faster. But then we see like you said like the general population, the people who don't use AI, they sort of get left behind. And I think it's a huge problem. We see that in like globally, you know, people who will not be able to differentiate deep fakes from from real images like but us, you know, who have a trained eye or know that that can't be right, we will know. But this whole thing we we have to sort of take everyone with us and make sure not to leave them behind. And I think with that that's always through education and transparency.

So I think education in terms of AI is is the most important thing. And it starts in every company. You know you have like AI um missionaries you know evangelists whatever you call them. And they sort of teach use case by use case like a little string of pearls and and teach people what to do to get them comfortable using it. And that's the other thing in in companies actually 85% of all project to pilot pilot to project fail because the organization cannot handle it and people don't don't adapt to it and it's simply because they they're not comfortable using it. So I think it's just like you know we have to all reach out and make sure people understand and it's mostly the older generation because the kids are they're already on board you know they they see the the use of it they see the beauty of it and they you don't have to worry about them.

Well, I I I don't agree. We still have to worry. We have to worry about not the use the usability. You're absolutely right. And it's really uh so that they learn how to use the things quickly. That's not the problem. We should still worry because um they just believe everything they see. There is no critical thinking. There's no critical thinking and uh they just take everything for granted and and use it. So still I would say this digital competence is something that young and old need different perspectives but uh uh yeah that's interesting because Ana you said we don't need excitement but we need probably literacy or how to use this tool or competencies what you just said and when other question do you agree that we leave a majority of society behind because they are not AI fluent?

Uh yes of course I mean I just um rely on numbers. I think we have an adoption rate in uh in Germany of around six to 7% across most industries. Um if it comes to the US I think we are already reaching up to 60% in China even a little bit more or a little bit less about the same. So of course yeah AI adoption means of course the adoption in a specific industry. So of course we are way behind. Yeah. And I don't want to disagree here. But what um I think is very often um a misconception is that especially when it comes to um companies and B2B applications, people assume that there are specific specific competences necessary or specific concepts in order to apply AI tools.

And that may be the case for a spec specific proportion of the tools. But the truth is there are many tools that can just be used. And I think that's an opinion that is said very rarely and that is not very um popular because of course people tend to um to to support the fact that first of all we need to develop a strategy and then we need to and so on and so forth. And I think the most important thing that harms innovation most is at the moment that we don't apply quick enough. And therefore, I'm just in favor of applying as soon as possible. And sometimes that very well works without any education before.

And that's just what I'm saying. I I totally agree with with with what was said. Yeah. Is it a problem of applying AI or is it a problem about culture and people? We already discussed what you said. Most of the I think it's culture and people. I mean, the tech is there. Yeah. and it's been there for decades and now now it's becoming actually very usable because we have the application layer on top we have a good UX UI on top so it's it's present but we we really I mean we have to work on on sort of like if you look into organizations we have to look like into change and if you look at societies we have to have like do sort of change management in a way you know to to get everybody involved but at the same time I mean AI will be woven into every bit of our society into every process into every software into every hardware like even my vacuum you know will eventually hopefully be able to

distinguish the cat from the chair so and have like you know enough mental power so I think you know it's it's just a matter of time I personally don't believe that you know like it's it's going to be tragic I think we just need to work on it you know what is how to work on it I mean and I mean not only talking about education but do you have examples how you get people on board when you do processes in your companies. Uh so something that we can apply when we go home. What does what is it that persuades people to use AI to find values to do this end to end digitalization to go into the kamu and then use AI?

How do you persuade them? What is what is what is the yeah trick? I think um in the workplace what is really helpful is when you make uh these people part of the things you build. Um and I mean obviously if you have thousands of of employees you can't include all of them. That's that's for sure. But um in my experience and we did this in different projects in the past even where in development solutions for uh uh um in the shop floor or or I don't know the call centers uh that was a very specific example we had workers from the call center joining the tech team from the beginning in this design phase.

Uh first of all be because they are the subject matter experts. So if we are improving their job, we need them on the table uh uh to know to understand their job first, understand the processes and then understand how to improve. And uh if you include them in this design process of the solution once it's built, they're really excited to use it because it's like their baby. And um not only they use it, but they talk about it with their with their peers. are like hey see this this automation piece I build it with the IT and it's my baby and so they they encourage the others to use it um so involving the end users in the design and development process of the solutions is really uh helpful in in the adoption and um I mean obviously offering trainings collaboration co-creation in in companies helps a But you need to do it in a structured way.

So it's all about doing it, experiencing it. So has you experienced it when you started in academia and then started with it? Would you agree? Um from my personal experience. Um so uh no I I think not actually. That's good. Something different is always good. Interesting. But I think I'm not also not that great at applying AI. I think I could be way better than I am. Yeah. Um I think I'm pretty good at developing AI, but um I think I guess there are many people here that are way better at applying AI to us than I am. Um yeah. So I got a little bit different excited about AI um which was rather through research and not through application tools.

Yeah. Yeah. What what I actually meant is the question do if we want to take the majority of society whether they are in companies or in education with us and also to um persuade them of the positive impacts of AI. Do people have to experience the positive impacts of AI? Yes, I think they so that's what you said cancer. Yeah, I mean of course if we take um the entire society as a whole I think definitely because AI is not just for AI. I think there should be a reason behind models. Um uh and I think um especially if it comes to companies there must be an economic value from applying a specific tool. If the tool does not meet specific KPIs then it's just not worth it.

And I think that's also what we need to accept that there need to be an additional economic value or an additional societal value. And I think as soon as people are realizing that that for instance let's take the example of um of let's take our example of a recruiting that you have better time to hires better retention rates or anything like this then it may make sense to apply AI in this field or to further develop or of course you need some time to get there but I think um AI is not something that is just there because it's AI it must have an economic value um yeah and I think that again motivates people yeah I think it's also also you have to you to differentiate a little bit what you know AI is like if we talk about models like we're helping you aiding you in writing an email so of course I think the the value prop is right there but you we have AI in everything in auto autonomous driving you

know we have it in in in medicine and and every every big chemical company right now is using um AI and and finding new materials there the the R&D process is shortened by by years and if not decades to find new molecule structures to find new um protein structures. So the whole like pharmaceutical industry will be disrupted. I think this like to me personally this like whole use case of of personal AI adoption is is boring in my opinion you know because we're all like I have my ChatGPT on the fast button on the iPhone on the right button. So instead of like making muting it and unmuting it, I have ChatGPT and ask it you know like whatever anything.

So that's my II adoption. I have four other models on my phone I use for various use cases whatever that's in my opinion super not interesting you know it's just text and image maybe. But what's interesting is if they find if they're able to find a vaccine against cancer, you know, 10 times faster, 10 years faster because they can like and cheaper because they can find molecule structures and then be able to roll it out to the entire like globe, you know, because there are systems predicting certain things, testing, you know, or for instance, um the latest models are now able to decipher animal language. So they found out that that pigs grunt have an entire like you know grunting system for their friends for other you know for foreign pigs and it's like once we understand all that you know what are the consequences for society if we understand what you know what a cow is saying before get slaughtered like is that going to change our our way we look at animals at

at livestock at at fe you know at at anything really so I think there's so many more interesting use cases like for instance in um in the ocean when when you have coral and coral reefs, the fish make a different bloop than when it's healthy versus when it's unhealthy. So Google deep mind the Google researchers went down and recorded that and as a training set they had actually bird voices like bird chirping that was supposedly similar. So they're now checking with little microphones down at the bottom of the ocean if if by by listening to the fish if the corals are healthy. And I think there are so many like just in science there's so many super exciting use cases obviously like on the company side you can create now marketing material on you know at the at a click basically you can create videos you can create images you can automate but that's all boring you know I think it's I think yeah boring maybe not very politically correct I'm sorry but let's let's

reframe it the value the value creation is at the bottom So if if we think of a of a triangle at the top we have the tools right and then you go AI systems APIs and then where the intellectual property is. So the value creation is going deeper in that in that triangle but mostly everybody is talking about the top which is tools I have I have peri and I I do the same like yes but there is no value creation in it. So especially if if we talk about the industry um we have to look deeper and in order to have real value to have real production uh gains we need the whole infrastructure we need the back end so the the front end the application layer okay that's easy and everything is plug and play today but what I I I always discuss very often with startups and we do work with with startups a lot they bring their solution It's ready.

It's plug and play. And they tell me, "Hey, it's really easy. Plug and play API and network." I say, "Yes, my problem is I don't have the plug where you can put your your solution. I don't have an API. I'm working with a system. My warehouse management system is based on an AS400. IBM very robust uh uh thing. Programming language RPG. Raise your hand if you know it. very old legacy. We have technical depths. The company is 151 years old. So, and this is this is the hidden champion situation basically, right? So, we do have a lot of homework to do in bringing up that infrastructure to a level where we can use the plug-and-play solutions and then the value creation starts.

Um, we have copilot. Yes, it helps for people to experiment to try but it's not really doing any positive impact on productivity. It's I still believe in having the um the workforce using these tools uh because I say on the long run if they experiment they will get new ideas in the future maybe in two years how to automate many activities tasks that they do daily. Maybe today we don't see the impact on productivity but I'm convinced we will see it in two years. Um so this is why I think still it's it's the investment is uh worth it and entire Germany is like this right so entire Germany has has their data in the basement in you know folders I mean entire Germany has legacy software I used to the the fund I used to work for we had software CRM software that was 22 years old and um certain people in the company worked on I want to say it was word 97 or so and that's just four five years ago so

I mean, and that's like serious legacy software. Like you there's no, you know, it's hard to believe, but it's hard to I mean, there was just no update forced, you know, when you work in big corporates, they all of a sudden in the call, they force an update on you and your laptop goes blank. Happens to me all the time. And um but there it doesn't, you know, like in and this big and and that's how like there's a lot of I would say opportunity in Germany. You can say it negatively. There's legacy, there's like tech debt, there's legacy software, legacy this and that. However, you can also say like the transformation that can happen is is really the potential is huge.

And also what's interesting is that they can um I've talked to a lot of uh health care systems and they they're really there lots like in the south of Germany they're really not digitized like really papers and like writing manual reports and and such. And the interesting thing is they can almost skip the entire digitization and go straight into AI. Yeah. It's like because that that all that like the middle stuff in the past 10 years they didn't have a CDO who tried to like you know vertically instead of horizontally vertically try to push sort of software and IT and new systems and maybe co-pilot but not even that but now they can sort of like try to have a horizontal approach where they like section by section business unit by business unit try to modernize it bring AI into every process and then really automate and I think so I I think Germany can on a good way if we do our homework at the end.

I'm still trying to figure out what is using your word the API is between our AI bubble and the normal the general public. Yeah. Yeah. Yeah. So and and these are the people in your companies that probably that that suddenly don't make the transformation work or stuff. What could the API for this be? I mean my answer was very analog writing books. Why is this? Well, yeah. And and and actually it's a question that I also asked myself back back then when did I write the first book 2020. Uh I was working at Telecom back then and we had this uh enablement sessions internally. We did a road show where we went to different locations and explain what is AI.

I I called those workshops AI for everyone, right? So just explaining very easily how the technology works and then uh um sharing information on use cases that we were applying in different in different business areas. And then I thought, yeah, and the reaction I had, you could really see on people's faces when they got into the workshop, they were afraid because, oh, AI, what's all of this about? And then after the workshop, like one hour later, they're like, oh, relieved and only because I explained it's just mathematics and some of this mathematics you even had in your studies, a regression. I mean, everybody heard something about linear regression. So, they were relieved after understanding there is no magical thing about it.

It's just applying applied mathematics. And then I thought okay if this happened within an hour let's say I need to share this capability with everybody. So this is what motivated me to write books for the general public. They're not uh uh science books but very very uh easily explained and that was the way I tried um uh to push that API let's say uh the connection that I that I found and of course there are different initiatives um different foundations belonging to big uh corporates also investing in education uh so actually the the there's enough offering out there and also for free. Uh, one of the initiatives that I really like is elements of AI.

It's a an online course you can do. It's for free that was developed by the University of Helsinki and they translated it in most European languages. So, the offering is there now is how to get the people to use it. But I very much liked what you just said that you explain people that it's in the end just statistics and that's basically it so people can somehow connect but I also think that sometimes um it may make sense also to talk about the weaknesses of AI and also to show how much AI is not greater than our minds are that it is actually artificial intelligence and not human intelligence. So that it's actually copying our code.

Um let's take Genai is just copying our code how we code our our human intelligence and then if you compare it you very quickly see that in terms of performance it's not so great especially if you perform you you compare let's say the output of Gen AI with a very excellent philosopher like Satra or something like that then you realize it's a way to go for AI and um then there are also funny examples I was just thinking about that I think it's called incident database or something like that. I don't know if you know this website where you can see AI mishaps. You see a vacuum cleaner going into a river or something like that and then you just realize it's not so great.

It's a lot of development work putting into that and it will be a way to go. So I I sometimes feel that if you see this then you you you you reduce a little bit the fear that many many people still have from you know general artificial intelligence or even super intelligence and all those scenarios that are drawn which are still so far away and when you see when you see exactly that happening or you could see exactly that happening with CLA because Clara announced I would say was it six months ago that they Um, oh yeah, they would get rid of 700 call center agents and replace them with AI agents and it was part of the equity story because they were planning an IPO and everybody was like in the industry everybody was super excited about it.

They made various headlines. It was a good good press coverage and I believe it was yesterday or so they recentered that and I said like well you know we might need to rehire a few people and and it shows like really the like the AI journey or the the problem everybody's in because we're not quite there yet like you said we are like actually far away in the call center space you will always need like a certain 20% or so to cover cases that have not been done before and then you will have like 80% of the standard use cases is like where is my account like can you give me up my number how long can I call whatever all the basics but the 20% is like a you will always have to have humans that's number one number two and that's just my assumption is that it's too expensive for cla to actually deploy their own models and record their own voices because otherwise you have an API call to 11 labs or

whoever voice provider you use and you have synthetic voices you need to pay you know royalty fees so every API you every time somebody calls the call the AI call center, you have on the other side, you know, a little dollar going ching because that's the API cool call to uh 11 Labs. And in order to prevent that from happening, um all the big companies like uh Lufanza, Deutsch, Deutschbang, everyone, they they started to record their own voices, create their own synthetic voices, but that's very expensive to train your own models, that's expensive, it needs to be open source, this and that. So CLA just now took it back, and I think it's symptomatic for where we are in the adoption process.

It's not going to be a linear situation. It's going to be like zigzagging. We're going to be zigzagging our way along along the adoption line. I think eventually it would level out a little bit, you know, but I agree with Anakah. I think these like these like maximum theories of AGI coming, ASI being around the corner and they actually propagated by by the um XTO of Google um or CTO. Uh they're like creating drama and fear, but I think we're so far away. So um I think people can take themselves can can you know take your time in in adopting and just uh move along. Okay cool. Maybe for the last question before maybe you have a question then after that one just think about it.

I want to zoom out again. AI and society. Yeah we talked about APIs how to connect but AI and society and we all know the doom scenarios you were just talking about. What is your most positive view on AI and society for the next three to five years? I mean, you already mentioned stuff like reading cow's mind and stuff like this. Um, what would that be if you zoom out? How positive how optimistic are you about AI in society? I think there there are um good or positive use cases if they are being used. So obviously healthcare is is is a very big impact but also on a communication level we have tools now that enable people to understand what's happening on television and you would say everybody understands but no not everybody understands you need to have higher education um so having those simple German translations and I guess in most languages translation between languages um courses online um there are also use cases to help children and also migrants to follow the school

curriculum at home in their own languages explaining everything. Um so there are there's lot of AI for social impact uh use cases and I think this also helps the participation the civil civil participation obviously there are back sides uh and and and we know that but we're focusing on the on the good parts uh in here Anika maybe um I mean I'm like a huge optimist on AI and I think especially I I talked in the beginning a little bit on the labor market and I mean that's a critical issue now for Germany. Yeah, we have we have actually many intelligent people, but considering that that um jobs changes and that job jobs change and that we have skills that are necessary for the future, we just need to realize that we have wrongly allocated people on specific positions.

We have people that are work less. That's not only because of sickness and all of those cases, but that's probably because people are wrongly allocated. And this is just one example. And with AI um with AI tools and AI systems, people are capable of really analyzing all elements that are relevant on doing a good job in the best way possible. So actually tools could in this case, let's take the example of the labor market, reallocate the entire German labor market in a new way. Uh which is of course a huge opportunity that won't be solved by our company alone, but will must be solved. Yeah. but by by the entire industry. Um, this is one case and also I mean one other obvious case that pretty much everyone is talking at the moment is probably the defense topic.

Yeah. the the the application of of adaptive systems and how much actually this entire topic and the the the principle of of defense industry and defense technology could be changed through good data and through adaptive systems that are working with real time information. I mean this is just a huge change in terms of defense situations that we have now in Europe which is more than more than critical I think at the moment. That's a very serious very positive advocate. I think I think I mean what are the greatest problems of humanity at the moment? We have like climate you know climate migration and resources. So we'll be fighting about you know at least we we be fighting about resources and I think AI can definitely solve that.

So I I think AI can can really um find us help us find new materials, help us find to be more efficient, help us to sort of like tackle the climate problem. So not the entire south of the globe will be like cooked, you know. So I think um more on a on a more macro level, AI will be magic for us. Thank you. I think at the same time just one one one addition, we shouldn't uh forget that we are in the driver's seat, right? At the end, this is just a tool and it's it's up to us to use this tool for a for a good purpose. We shouldn't we shouldn't forget that because sometimes you have the feeling people just lay back and say, "Okay, AI is fixing it."

No, no, no, no. We need to fix it. Uh so yeah, but maybe to add something here, I have never met any AI developer who does not want to work for a good cause. Really, everyone wants to do something good with it. And that's a nice uh thing. That's a nice words to the end of our discussion. Do you have any questions? No. So my last very last question would be AI society. Uh how optimistic are you? 10 is very optimistic. Zero is not optimistic. Would you give a number KPIs? 100 of course 100. Ah, I'm skeptical about this question because as I said, fair enough. AI is just a tool. Okay. And and uh we need the right people to use the tool for the right purpose.

So this is why the question is not really answering AI as a tool is capable of many good things and bad things. But the question is how do we use it? And people are very creative and surprising sometimes in both ways. So I'm not so sure. isn't it? Me too. I would say I'm about a five and I just really quick yesterday I read that the Russians are sort of spamming the internet with false information. No news but it's now being straightly like uh fed into ChatGPT. So PUFDA I think that's one of the big Russian media houses um has managed that Chad TPT uh generates wrong answers about Russia in 30% of the cases.

And so I would say I'm at about five because there are too many bad actors and it also amplifies their skills you know. Okay. So we have a five we have a it depends on the people and we have and we have a 100. Thank you very much. Thanks for this panel. Dish. Thank you.

Rise of AI 2027 · Berlin · May 2027

Be in the room for the next one.

Ten years of talks started as conversations in one curated room in Berlin. The next edition brings 300 selected leaders together in May 2027.