Still Human: Real Talk in the Age of AI
AI is rewriting how we work, how we build, who benefits, and who gets left behind in the process. We've been told AI is the great equalizer. The data says otherwise.
Still Human: Real Talk in the Age of AI is the podcast that goes inside the rooms where the future is being built and asks the questions nobody's asking on stage and puts the raw conversations on record.
Host Nadia Koski is a marketing and brand strategist as well as podcast network producer, with over a decade in media, content innovation, and advertising technology. Every other week she sits down with the builders, leaders and thinkers navigating this moment: CEOs and startup founders, data scientists and creative directors, NGO leaders, engineers and entrepreneurs doing remarkable things across every time zone.
The conversations are candid, unscripted and go behind-the-scenes on what we're really doing with AI. Nadia gets under every guest's skin, the mental blocks, the habits, the stories they had to unlearn to get where they are today. The question underneath all of it: what does it actually take to stay human while the world rewrites itself around you?
Still Human: Real Talk in the Age of AI
Live from OMR, Hamburg: What Can AI Never Replace?
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
Seven conversations. Two days. One question running through all of them: What does it actually take to stay human while the world rewrites itself around you?
Recorded live at the OMR Festival in Hamburg, Germany, this first episode of Still Human brings together seven leaders at the sharp end of the AI conversation. Stefanie Tannrath, CEO of Universal McCann Germany, on why our entire frame around AI is too small. Yulia Kalner, Head of Media & Creative at Kantar Germany, on the hard data showing AI makes brands more efficient — but not more effective. Sabine Mayer, Chief Learning Officer at treibhaus, Marketing & Learning Lead at Take-Aw.ai, on the skill nobody's teaching yet: the one that sits somewhere between resilience and reflection. Christina Sontheim-Leven, CEO of 3e, executive mentor, and co-author of "Machtgebiete", on what genuine creativity actually is — and why AI is still just remixing what already exists.
Also in this episode: Marija Lux, Director of Customer Success Central EMEA at Integral Ad Science, on why prompting an LLM is nothing like a Google search — and what happens when teams share their AI experiments across departments. Rosa Markarian, independent consultant for AdTech, AI, and sustainable tech, on the market knowledge, relationships, and intercultural fluency that no model can localize — and on the energy cost of AI that nobody wants to put on the agenda. And Michael Mörs, Founder and CEO of SMA Development, who has been building AI into real products since 2017, on why you should never trust the output too much — and why the current wave of "AI-driven" layoffs is better storytelling than causal fact.
Thank you to MMA DACH for their support and partnership on this episode.
Still Human is available wherever you listen to podcasts. New episodes out every two weeks. Please follow, subscribe and leave a review if this one landed.
Sound engineer: Morgan Bosc
Cover art: Adam Richarson
Nadia Headshot: Alessandro Fibbi
Intro
SPEAKER_05Hello, hello, and welcome to the very first episode of Still Human Real Talk in the Age of AI. I'm your host, Nadia Koski, and I am so glad you're here. This episode was recorded live on the ground in the middle of one of the loudest, most important, and most exciting marketing and tech conferences in Europe. OMR, Online Marketing Rock Stars. This took place in Hamburg, Germany on May 5th and 6th. This is a conference where 70,000 people swarm the grounds to meet up and talk tech advertising, you name it. There were hundreds of stages, Porsches turned into bumper cars, and one word being shouted, whispered, debated, dissected in every single conversation. You guessed it. AI. So I did what makes sense. I grabbed my mic, I found a semi-quiet corner on both days, and asked seven brilliant leaders the questions bouncing around in my head. Not just the fluff or the AI is the future sound bites, blah blah blah, but the real talk. We dive into what's actually working, what's not working, and what are we still missing from this conversation? You're about to hear from seven extraordinary humans, a CEO who'll tell you why our obsession with efficiency is the most short-sighted view of AI on the market, a global market researcher who'll explain why consumers reject brands that put AI in their face, a chief learning officer building the next generation of live events, a best-selling author taking on the system holding women managers back, to the tune of 372 billion euros a year in Germany alone. A customer success leader who will convince you to use AI before your second coffee, and a go-to-market leader who refuses to talk about AI without talking about its energy footprint. I love those opinions from Rosa. And the founder of a company that transcribes 100,000 podcast episodes a month with some of his tools and his uh software he's built, and has some very interesting, unfiltered takes on the layoffs, the hallucinations, and why coding is the easy part. So I asked them the things I really wanted to know, which included what's the most dangerous assumption we're still making about AI, what can it never replace? Uh, what's missing from the conversation we're already having? What has surprised you about AI for the better or the worst in the last six months? Maybe what you would say to someone that's still hesitating about using it. And the other question that always kept coming back to was how are we winning with AI? What does that look like? And what's that gonna look like five years from now? Because the numbers underneath these questions are quite loud. 78% of organizations now use AI in at least one business function. Men are 22% more likely than women to use AI daily at work, and women's jobs are three times more likely to be automated by AI. So, right now we're really talking about how we're going to be shaping the next decade. And who's gonna have a seat at the table? We can't fix what we don't talk about. So we're yapping up a storm. A big thank you, Dankushun, to MMA Doc, for partnering with me on this episode. Your support was incredible. Grab your coffee, grab your walking shoes, or whatever helps you think, and let's get into it. Real talk, real people, still human. Andiamo.
Stefanie Tannrath, CEO Universal McCann Germany
SPEAKER_05First up, Stephanie Tanrath, CEO of Universal McCann Germany. She walks straight into our biggest blind spot. Why we keep talking about AI like it's just a productivity tool, and what we're missing when we do.
SPEAKER_06We're live at OMR. I'm here with Stephanie Tanrath, CEO of Universal McCann. And it is an exciting, bustling day. Hello, Stephanie. Hi, yes, it is. Thanks so much for having me. Thank you. Um, well, one thing that everyone is talking about at OMR is what, in your opinion? What's the buzzword?
SPEAKER_10The buzzword, it's it's a very simple answer is AI, obviously. And it's at the at the verge of, oh god, I cannot hear it anymore. But um And that's why we're gonna talk about it right now.
SPEAKER_06That's why we're gonna talk about it now. Um well, in your opinion, what's the most dangerous assumption that people are still making about AI right now?
SPEAKER_10I think a lot of discussions about AI are relatively short-sighted because it's just about efficiency, it's about how to increase productivity with AI. How can we uh save time, money, resources? You know, it's it's all about uh making things um faster and uh more productive ultimately. But I think the the bigger picture is there's so much more to it because it will change basically everything. It will change how the world works, how business works, it will change job profiles, there will be uh less jobs potentially, but different jobs, completely new ways of working that we cannot even maybe foresee now, or at least I can't. Um so I think you have to think it in a much broader sense, but also beyond business. So when you think about you know how the way how we see explore the world changes so much with AI, it already has now such a big effect. And um I think yeah, that maybe is sometimes uh a little bit of short-sighted view that we have on it.
SPEAKER_06Yeah, and it's going to change a lot every every different sector, not just media and asset. Exactly, exactly. And especially healthcare, which I'm very interested to talk to in future episodes, some people in that field, because what I read, especially in finance and healthcare, where there's more sensitive data, AI is only really being used up to like 8% of its like potential in those fields. And so there's a lot of room for advancement um going forward.
SPEAKER_10Absolutely, and especially in those fields, I think it will totally change the way we see healthcare, etc., because it's obvious it's so easy for you know first-level care kind of uh situation when it's not about uh life and death situations, but it's just you know, you and how whether it's immersion and also mental health, all sorts of things, but yeah, very good.
SPEAKER_06And I think what we're seeing also a lot of companies there may be layoffs, it's happening, it's inevitable. And a lot of people think that AI can do the job, and there's maybe not a process in place. So uh that's one of the things that we'll be talking about. It can lion, um little zinger there, uh little prequill. But uh yeah, completely agree with you. And we have to understand better how those processes are working and not just be more productive and faster, right? Exactly. Yeah, exactly.
SPEAKER_10And also how it will not only we all we always talk about how we will change our work life, but it will also change um job descriptions, and you know, there will be new jobs created through AI as well, and we have to also consider that.
SPEAKER_06Yeah. Well, what else would be missing, in your opinion, from the AI conversation right now?
SPEAKER_10I think what there's probably lots of things missing, but what we're missing is definitely um marginalized voices. Um I think that's that's a big one because obviously with every big tech, it's the the the people behind the people programming it, the people you know creating the code are the ones defining also then how the AI thinks, and then yeah, thus how we think. There's gender bias, there's definitely a gender bias. Um that's that I think that's a that's a big danger, especially in the times we live in now. But also all other kinds of marginalized voices which don't have um the the the gravitas maybe in that conversation or in the creation of code of AI, etc. etc. Yeah. And I think that's something that um we should all also as a society consider when talking about AI.
SPEAKER_06I love how you touch on that because uh I'm super nerdy and go down rabbit holes when a topic excites me. So I've been reading a lot of reports recently. I was also a UN uh women delegate at the CSW in New York in March, and this was a big topic of discussion moving forward with these new technologies, having digital safety for women and girls, and then having that representation of everything. And one of the um crazy stats and a fact that's from the World Bank is that only 56% of data is actually being used to train AI, which means AI is training on half of the world's missing data. Yeah.
SPEAKER_10So which is wild.
SPEAKER_06Isn't that yeah, scary as well? Yeah. So there's a lot of improvements we can do. Um, to close out, what is one thing that you can't wait to do right now at OMR in the next two days?
SPEAKER_10Oh, I just got I can't wait to have like to meet lots of people. It's a little bit like you know, a class reunion always. Yeah, so I'm looking forward to that first glass of wine later and just have really nice and good conversations about AI potentially.
SPEAKER_06The buzzword that no one is everyone's talking about that no one wants to talk about anymore. Well, thank you so much, Stephanie, for coming on and taking the appreciate it.
SPEAKER_05Next, I spoke with
Yulia Kalner, Head of Media & Creative Kantar Germany
SPEAKER_05Julia Kalner, head of media and creative at Kantar Germany. She brings the consumer data the rest of us don't see and lays out the big gap between AI being efficient and AI being effective.
SPEAKER_06We're live at OMR. I'm here with Julia Kellner, who is Head of Media and Creative at Kantar Germany. Amazing. And this is a wonderful opportunity. I'm super psyched to talk to you because you're doing a Kantor does a lot of research and a lot of looking at what's going on between the marketing and and the brands. So we'll get into that in a second. But at OMR, uh it's it's fun, it's overwhelming. It's your first time, it's my first time. Um what is one thing that you really love that you're gonna take away and take back with you? I may miss.
SPEAKER_02I was impressed actually by, first of all, people here, I have to say. And this is my biggest impression from the OMR, regardless of how many different conversations we have here, in the center of this and at the heart of this are people who are making this industry and just looking around and seeing these wonderful people, meeting colleagues, being excited from what they share here, is kind of my biggest impression from the OMR, and I'm really impressed by people who are here. So it's one of the greatest events I ever had in Germany. So I like it.
SPEAKER_06Oh listen to that OMR. Uh yeah, no, I I really loved it. Um, and I there's so much information that I suffer a little bit of FOMO because I'm like, I have to go here. No, I gotta go there. I gotta go masterclass.
SPEAKER_02I kind of accepted this from the beginning, and I was like, okay, I will have a FOMO. So that is why I can it's fine.
SPEAKER_06So good, good. But we're here. Um, and this is going to be officially the first episode of the Still Human podcast, and we're all about the real talk, what you're doing with AI, how is it changing, what's working, what's not. Um, so it's it's a great opportunity to talk to you. And there's been a lot of development, even in the past six months, let's say. I mean, Claude Cowork has like blown open the doors, so many things happening. What does that look like that you're seeing at Cantar for brands and marketers from a Germany perspective and then also wider global perspective?
SPEAKER_02In terms of the AI I mean, I mean now yeah, it's all about me, what is going on? I have to say it's first of all with AI, and for sure, kind of what we do with AI is kind of on two pieces. So, first of all, for sure we use AI in our day-to-day work, and it's really kind of the thing that helps us to be more efficient, faster, get everyone's using it more efficient, more efficiently. This is exactly the point. But also, another part of me working with AI is kind of talking to brands how AI is actually helping them and specifically with the communication. And then I kind of send a point of the consumer because as a market researcher, you introduced me as someone who is kind of in between of kind of explaining better on the insights, but I actually the one who is explaining the brands consumers and their point of view. Yes, and this is kind of the thing, consumers do not really like what brands are now doing with AI. They struggle with this, and even though What don't they like? What don't what is it that's uh they're not I mean, when AI is kind of into your face, it's something then consumers start rejecting. And this is like, my god, no, that's not for me. So they're not against this, but the level of kind of excitement about EI across marketers is that high. The level of excitement about EI across consumers is so low, just because they're like, oh my god, I'm scary. And specifically in Germany, so I probably will start talking stereotypes right now. But Germans do love, understand how things are done. And specifically with AI, you do not really understand how things are done. So what how being produced and stuff? So this is this is not that is why, yes, you are right, everyone is talking about AI helps me to be effective. The problem is efficient, sorry, you said AI has to be efficient, yeah. But the problem is AI doesn't really help us yet to be effective. And there is a difference, a big difference between being effective and being efficient. Yes.
SPEAKER_06AI. 100%. Uh huge difference. And we're here and we're here. Yeah. And what I've been hearing also in other conversations is um we're using it to do things faster, not necessarily better, sometimes better. Um, and then that saved us time. We have three hours of extra time. Did we go for a walk or play with our kids or you know, do some do a painting? No, we went on to the next task, so we saved that time and we're just getting faster, faster, faster. Yeah, and that's something that I I think about and try to be more conscious of because it's like, okay, well, let's let's slow down a second because this is also a technology that's kind of making us move constantly faster, which is Yeah, and interesting kind of the stress level.
SPEAKER_02I I really agree with you, because you're kind of yeah, it's at some point it's kind of at the speed that you cannot actually follow anymore.
SPEAKER_06Yeah, but what's uh going back to how consumers perceive it and how marketers are super excited about it, um really interesting because I know you had research about this, and uh I was looking at some things when I gave a talk last December, so just a few months ago, um, and I was trying to understand and explain what is a brand brand perception, how is that built when people are looking at AI? And I had uh a spot from Coca-Cola that was all AI, and then a spot that was the um the new iPhone spot that was all like uh old school Muppeteering. And and it was interesting to see in the audience like how they also reacted to it. What what what is it that like especially in the creativity area? Like, is there something that when people see AI, they're just like meh, what what what is it that turns them off?
SPEAKER_02I think that we as marketers kind of overestimate what consumers think and do and discuss about advertising. So even when you say, I've seen a lot of conversations about these two sports, that's not what is happening in real life with consumers.
SPEAKER_12It's what we marketers discuss on LinkedIn. Yeah, exactly. But that's not what consumers actually discuss.
SPEAKER_02Consumers like good advertising, they can discuss it, but they will never discuss how it's produced, whether it's AI or not AI. They discuss emotions, what actually what actually stands behind this advertising. And that it's our job as marketers to kind of produce ads that generate these emotions with the help of AI or without the help of EI. That's like a technical it is behind this, but for consumers, it doesn't really matter. We just don't need to produce something that they kind of will reject. And when we put EI into their face, that is something that they reject right now. Can give you a short example. Yeah. And kind of with the EI, we all believe EI knows how to do things right, and specifically kind of technical things in advertising, like make sure that the brand is visible. This is kind of the technical thing quite often, and the eye should know at least this one. So, probably emotions, it's a difficult branding, it should be easy. But what we know from our database, for example, that branding for AI-produced ads is actually even lower than uh for the ads that are produced by humans. Which is kind of a funny fact, just because I mean you would believe that AI should know kind of the basics, but even with this, uh you still need kind of a human. For sure it will develop and AI will learn its stuff. But right now it's not any better.
SPEAKER_06So that is super interesting and a good point because we're always ready to say how it's great, how it's you know, a billion brains working for you, but it's still that is why that is why I kind of keep saying to marketers.
SPEAKER_02Yes, it's true, it keeps it it helps you to be effective. But I'm not sure that it helps you to be efficient. Uh otherwise, it'd be efficient, but not necessarily.
SPEAKER_06Exactly. Yeah. Exactly. So since all of there's a lot of development, there's so much change going on, and actually, what we will explore more on the show is um specifically women, how we're using it and why we have to be using it uh in certain ways, because it is going to change every field, not just media and tech, it's going to uh you know, it's in finance, it's in healthcare. And what does that look like? So for you in two or three or five years, let's maybe in the more media. What does it look like winning with AI? And what do we have to do to get there in your opinion?
SPEAKER_02I mean, there are different things, but I think the two most important ones are first of all, that the data that is being used for AI to learn, we quite often forget about the quality of data that stands behind this. We kind of trust AI that it learns based on kind of everything, and this everything is kind of quality data that we trust. I'm not sure about this, and I think that we still need to make a bit more thinking about what is kind of the input we give into machine to make sure that we trust the outcome. Yes. And by this input, I mean the data quality itself. And another thing is kind of in five years for sure, the ones who will win are the ones who can use AI in the form not to be generalists, because the biggest issue of EI is kind of and the biggest feed that the the the strongest feedback that I hear from my clients, AI doesn't allow you to differentiate. AI gives you quickly ideas that are really general. Yeah. It doesn't help you to differentiate. But we know that brands who win and who grow are actually the brands who are able to be different and meaningfully different to multiple.
SPEAKER_06And have their personality and their human inside come out and have that emotional connection and not have a generic you see you hear the term AI slap, just it that really does. It kind of generalizes everyone. So it's the individual voice that comes out. Yeah, exactly. Well, thank you so much. This has been a lovely chat. And uh I know you're on your way flying out, so take care. And it was such a pleasure talking to you.
SPEAKER_12Thank you for inviting. It was a pleasure to meet you as well. Yeah, thanks.
Sabine Mayer, CLO treibhaus, Marketing & Learning Lead at Take-Aw.ai
SPEAKER_05Now, Chief Learning Officer at Tribe House and marketing and learning lead at takeaway.ai. She's training the next generation of the events industry, and she's got a phrase you're going to want to steal resilience and reflection, people.
SPEAKER_06We're live at OMR, and I am standing here right now with Sabine Meyer. And uh, Sabina, you are, I love this, CLO, which is Chief Learning Officer at Tribe House and Marketing and Leaning uh Marketing and Learning Lead at Takeaway. Welcome. Hello. So nice to meet you. Thanks for having me, Natya. Thank you. Uh what I wanted to understand first and foremost here at OMR is uh there's a lot going on. It's it's fun, it's chaotic, a little overwhelming. It's my first time. Um but uh since we're all human and it's about the connection, is there one particular conversation that you've had that's really stuck out in your brain?
SPEAKER_11Absolutely. So I had one conversation this morning because we went for a pioneer's breakfast. So we had our own uh space where we met with around about 12 C-level persons from the event industry. And I had one conversation about AI, and at the moment we discussed about what to do, how we could get more efficient, efficient, efficient. But let's start thinking about the things we can't do right now. Yeah. So let's really try to think of connections, content, ideas, human constellations, whatever, which are not possible. Possible at the moment.
SPEAKER_09Yeah.
SPEAKER_11And then maybe we could find a solution based on AI or in context with AI, instead of yeah, only talking about efficiency is the answer and not asking what was the question. And then understanding is efficiency the answer?
SPEAKER_06Yeah, I love that because I feel like we've been in the AI trenches now for a couple years, and we've seen it just really the use and what you can do with AI in general, authentic AI moving so fast in the last even three, four months, right? Um, and it's uh it's wild how you can almost forget. Let's use this and keep the human in the loop, understand how we can connect better, um, and and not we have some rolling, rolling friends coming by, and how how it can improve and not just be a productivity to make us go faster and do more, right? Um, but speaking of, uh, is there something that AI has done or surprised you with in your business in the last six months?
SPEAKER_11Just now. So I remember especially one workshop which I had at the George P. Johnson with a very young uh colleague, I think 23 years old, and we used Hicksfield as an AI platform. And the quality of the multi-prompting output, so doing everything together, creating video, creating voiceover, creating sound, and within 15 minutes, the whole typehouse team produced something like advertising spots. Yeah, and that's it was really surprising to me. Wild, absolutely wild. But at the same time, of course, this is connected to efficiency again. Now we can do it so fast. Of course, we have to check what is this, what what is the product we created. What's the outcome? How do people react to it? It's a playful way to get into the tool or into the platform and understand it's not about knowing this tool and this tool and this tool. It's uh I think most interesting is to really compare tools at the same time and then get a deeper understanding of what really helps me for my special project.
SPEAKER_06And what you like, and what's uh what what is giving an outcome that is, let's say, up to your level of of I don't know, uh the emotion that you wanted to create, the the look and feel of it, if it's if that's the right thing, or is it just maybe slightly a little off? Or I I working in video so many years, like it is crazy that you can just put in a very descriptive prompt and come out with a video in like five minutes. But is that going to connect with your audience? That's what I always think about as well. Is that going to create some type of emotional connection? Yeah, um, as opposed to having maybe your crew and and uh a team of very professional people doing a specific role and all that to create something instead of just poof, it's here.
SPEAKER_11Yeah, yeah. So the the output is a little bit more random because you don't think about every scene and every light and all the details in depth. So, but I think it will change the way of thinking about the output and maybe not plan so strict each millisecond and give the random factor a choice.
SPEAKER_06Yeah, maybe, yeah. But yeah, that's it's super cool what you can do with um with a lot of the AI for for video, for voice, for creating things. But um, let's talk a little bit about uh takeaway. And so what takeaway does is it turns audience thoughts into slides in real time. So in an event space, for example, what happens when people see their their words reflected in like immediate real time? It's on the screen. Is there what kind of reaction do you think people have?
SPEAKER_11So our tool, so we have no competitor at the moment. That's why whenever takeaway is used, people are like, wow, that's possible. Because it's like you have these thoughts in your brain, and we always ask the audience. So when they went for a keynote or workshop or an event format, and there was interesting input. And for example, here at OMR, you jump from one session to the other, talk to people so many things at the same time, yeah, and two hours later you can't really remember what it was. What was the inside of this very interesting thing? Yeah, what was the takeaway that I wanted to? Damn, I didn't take a note.
SPEAKER_07Yeah.
SPEAKER_11And with takeaway, you create this moment of reflection that you really have the time, and just two minutes is enough. It's enough to go for this thinking and speaking at the same time. And of course, the magic is that all the people do it together, and then you see it on results, so written words in real time, and it's like, wow, okay. Now I understand most of the people are on the same track, and maybe ah, this point I didn't even think about.
SPEAKER_06Yeah, that's a nice collective way to absolutely get all the information together.
SPEAKER_11Yeah, and it creates a feeling of we all spoke together with each other, even if we spoke to our phones, but we had the chance to put it all on stage and creates this collective moment.
SPEAKER_06And that's what's so cool about the in-person events, too, which are making a comeback, I think, because a lot of people are working remotely, I work remotely, and so having connection again and the collectivity and that exchange is so crucial, I think, to how we're moving forward. But um, you're also basically building the next generation of events um with with Tribehouse and using AI to reshape how audiences uh participate, basically. But what is a skill that you're maybe most urgently trying to teach people that didn't exist five years ago?
SPEAKER_11Yeah, so it's a quite interesting question because what I realize at the moment is that companies, the event industry has a challenge to do things even faster with lower budgets, and the creatives are have the challenge to really I have two hours, I have to do this, then I have to produce the ideas, and on the other side is a client, maybe she or he prompted a picture and has something in mind, and you're so under pressure. But we tend to do many, many things in fast pace, but sometimes we forget to ask are those things the right things. Yeah. So what I try to teach is a mixture of resilience and reflection.
SPEAKER_06Maybe we could build a new word of these two I need a new word in English and in German for this because that's very true. We're so caught up in efficiency, productivity. Yeah. But is how is it going to be yeah, the the resilience or the how is that the right thing that we want?
SPEAKER_11Yeah. So just think about this maybe two hours time slot and then take your time to realize what I can do with this time slot and what is the output for now and for the next step. And not always think about oh, I have to do so many things at the same time. So small packages will help, huh?
SPEAKER_06It's it's like uh and this is uh an interesting conversation that um also people have been talking about online is AI can make us faster and more efficient, more productive, but are we actually saving time? If we just save three hours, we fill that three hours with the next thing to do instead of okay, well, I just got three hours back. I can do something else that's going to relax me, or I can go play with my kid, or I can go for a walk in the park, or are people doing it? I don't know. They're going on to the next thing. Yeah. Um, so and that's a little bit scary because it just means going faster and faster and faster and faster and faster. Um, but yeah, it's it's something to think about, and the reflection part is absolutely on point because we we do have to kind of think about how and what did I just create, how did I save that time and what I'm gonna do next with it, and and not save you the next task.
SPEAKER_11And I think on top, the younger generation can help us saving our time to do all these beautiful things like going for a walk or whatever, because they really try to focus more on what is my working time and what is my yeah, leisure time. We can learn from them. Often so, we we we are something like front it, but I think we can learn from each other, and on top of this working culture, those are the experts for AI because they start working, and for them, I AI existed all the time. Yeah, yeah.
SPEAKER_06For us, it's normal, yeah. Yeah, yeah. Um, so last question to wrap up. Um, speaking of you know, working in live events and learning and AI, which I think is such a cool combination of things, um they all require that human presence and that connectivity we're talking about. So, what's the thing that AI absolutely cannot do in this space right now? Um, no matter how good or how advanced it gets.
SPEAKER_11So, this morning I had an interesting um conversation about the definition of intelligence because AI seems to be a second brain for us, but this is not the only intelligence we have. And I like the wording live intelligence, and I like thinking about what could this mean, because at a live event, all those experts work together. Experts for media, for the stage, for catering, for content, whatever. And an event like this could never happen if you haven't had all these experts working very close and very human together. So I think this is one factor of live intelligence, and the other one is collecting all this brain power in the heads of real humans. And yeah, make them visible. Talk about it, not even yeah, we all walk here between very interesting people and manage to talk to just maybe 20 of them. Yeah. So, how could we connect each other, really strong in this life moment through life intelligence? This is one question which I really want to think about in the next weeks.
SPEAKER_06Yeah, I love that. Yeah, well, Sabina, it's been wonderful chatting with you. I love all of the interesting points that we we touched on. And uh, thanks again and have a wonderful second day here at OMR. Thank you. Thanks for having me. Yeah.
Christina Sontheim-Leven, CEO 3e, Executive Mentor, Co-author "Machtgebiete"
SPEAKER_05Up next, Christina Santheim Levin, executive mentor and co-author of Macht Gebeite, a new book on what female managers experience and how they push back. She makes the case for ethical boundaries in AI and for changing the system, not the women. Thank you, Christina.
SPEAKER_06We're here at OMR, we're live, and I am so excited to be speaking to Christina Santan-Levin. And we are here, and this is an amazing fair, as always. Um, what is uh your experience uh so far at the fair? What is something that you've been excited about?
SPEAKER_01Well, it's it's like a threefold thing, like you know, Kinder überraschung because you have, of course, you have the festival character, uh, which is I mean, you're popping into an astronaut, you're you're you're you're having that auto scooter with Porsches and stuff. It's it's a huge playground. Yeah. But it's also the most inspiring place to be and to look at the predictions for what's coming up next. Yeah. And then of course, it's networking, it's networking powers. I mean, here 70,000 people coming together, and that's a huge power cluster.
SPEAKER_06Yeah. Um, well, Christina, you're the CEO of 3E and also an executive mentor. So I'm really excited to get your take on some of these hot topics. But first, is there an interesting uh or something that's in every room that everyone's talking about that you've noticed?
SPEAKER_01Yeah, yeah. So the one thing you hear in every room, in every masterclass, in every keynote is of course the agentic AI. So, really, last year everyone was talking about AI answering questions. Yeah. Now everyone is talking about agents really pursuing projects. So you have masterclasses like um build your own shopping agent. You're talking about, okay, where are the biases in the agents? How do we give agents really tasks and do not give them our own prejudices? Um, you talk about um the danger of agents, you talk about the bottleneck that are really existing in organizations when using agents and so on. So, really, I think that's the topic.
SPEAKER_06Yeah, and I love we're gonna break that down a little bit more because on Still Human, that is the podcast where we go into detail about what is actually working, what is not, what are you using? I don't want the fluffy overall, I want the details. Yeah. Um, so also as I'm sure as an executive mentor and working with humans so much, but now we have to work with AI. What's in your opinion something that AI can never replace?
SPEAKER_01Something that we can still bring in as humans is the the genuine struct, to be honest, because what AI is producing is to a certain intent creative, yes, but it's reproduction of something that is already there and just mixing it up again, but something really new, a really good and new idea, something that is totally out of the box. I think that's still something that we can bring in. We can be supported by AI to redefine, to distribute, and so on. But I think really creating something new is still something that we can do better than AI.
SPEAKER_06Yeah, well, human creativity is at the core, right? And AI is helping us maybe with our idea creation and the process and coming up with it. But the human in the loop is is the actual.
SPEAKER_01And by the way, I mean uh in in the predictions yesterday from Scott Galloway, yeah, he said like the human and AI companionship will raise. So our best friend will be maybe um the AI agent. I don't know. I don't know how I feel about that. No, no, no, me too. And and honestly, I I would fear if that would happen, yeah. So I think uh that's another thing that we as humans can still do. We can be really compassionate, not like flattering the AI. I mean, if you talk to an AI say, great idea. Oh, you're so brilliant. Reinforce it. I can feel you and so on. But it's different if you have a real human on the other side saying, yeah, I feel your pain, I feel your thoughts, whatever, your fears.
SPEAKER_06Yeah. Also, and I'm here to support you, and how can I help you?
SPEAKER_01That's maybe one thing that's were you within the real life. I mean, sitting together with a real cup of tea is different than being with your phone on the sofa.
SPEAKER_06Yeah, yeah. Uh it's it's a really interesting discussion, especially for young, younger generations that are now using the tool. Yeah. And they have to almost be reminded and learn that human conversation interaction and and not get caught up in the AI competitive.
SPEAKER_01It's it's the COVID generation. I mean, they were they were forced to learn different ways of interacting, and now they have to relearn social real interaction.
SPEAKER_06Yeah, absolutely. Well, um, another thing that I wanted to get into was is there uh something that is maybe missing from our AI or authentic AI conversation right now?
SPEAKER_01Yes. Um, from my point of view, we're not talking often enough about the ethical point of AI. So, what is our obligation as leaders and as humans, by the way, to think about the ethic boundaries that we give to AI? Of course, we we talk about the threats like okay, AI taking over world control. Yeah, but between like, okay, we use AI for shopping and taking over world control, there is a between. And and but for this in between, we have to create ethical boundaries. Every company using AI has to think about where are our red lines, where we will refuse to use AI in our decision-making process, in our design process, in our interacting with our employees or our consumers. Yeah, and this is a point that is not addressed as far as I see a lot.
SPEAKER_06Yeah. Is there something that could help do that better and help companies think about that ethical side?
SPEAKER_01Uh put put it intentionally on the agenda. I'm, for example, I'm working with a big law firm together, and they have been inviting me to talk with their top leaders just now about this topic. So they're really taking the time. Um, and I think this is important to make this first step and be conscious about it. Yeah, yeah, absolutely.
SPEAKER_06Well, Christina, you're also a co-author. A recent book just came out. Let's talk a little bit about that. Um, first, you say it in your beautiful German natural language. What is the title of the book?
SPEAKER_01Well, surprisingly, I have that book with me. So it's called Machtgebiete. Was manager and wie sie gegenhalten.
SPEAKER_06Amazing. And in English, that would be Power Territories, what female managers experience and how they counter uh counteract. So uh this is a passion topic of mine, is um formerly talking a lot about women in leadership in a previous show that I did. Um, and you have a tagline that is change the system, not the women. Oh yes. I love that. Yeah, because let's be honest, we're not the problem.
SPEAKER_01It's the system that's the problem. Tell me more. True. Well, I mean, we've been spoken to 50 top manager, female top managers in Germany, and it turned out they all encounter the same behavior. They all encounter the men's veining, the interruption, the ideas that are uh stolen from them, the the microaggressions and the big Me Too cases, even if they are in boardrooms. And it turns out uh from this, yeah, it's not the women that we have to change, but it is a system that we created, bringing those brilliant women not into their full force for them, for the companies they're working with, but also for society. And this is a big loss. And we have put a number on that. So it's 372 billion euros per year that we are losing for the whole economy, just for Germany, because we don't let them participate on the labor market as equal as men. So, yes, and and this is why we don't we don't give hints like how to be a copy of a man. No, we give hints if you are part of the system and we are all part of the system, where in your circle of influence can you help to change the system for the good of all of us? Yeah. Men and women.
SPEAKER_06Yeah, and it's it's it's really about having the discussion together because that's another big point that has come up in topics and conversations. It's like um the system is not really uh adapted for women, but we have to have a conversation, men and women together. Now we can fix it. Yeah, that's essential. It's not just it's not just us talking and talking, talking with each other, talking and talking.
SPEAKER_01And I mean, yes, we are 50% of the population, but we need the other 50% because only if we work at 100% in the same direction, yes, we can have the system to be changed. And by the way, we we will all profit from it, men and women, our kids, everyone, everyone. Everyone, yeah.
SPEAKER_06Well, I'm so excited. I will maybe attempt to read it in German with a translating tool to help me with the English. But I'm so excited. This is such a great um book that it's it's out and it's available. So thank you so much for taking the time and having this chat. Uh it was great and enjoy Omar. I will. It's my first time here. Oh my goodness. It's my first time here it's mind-blowing, isn't it? Yeah, it is. It's it's amazing, a little bit overwhelming, but so fun. So it's yeah. Great. Day two, here we go. Thank you so much. Thank you. Bye.
Marija Lux, Director Customer Success Central EMEA, Integral Ad Science
SPEAKER_05Now up, Maria Luks, Director of Customer Success, Central EMEA at Integral Ad Science. Her rule a good AI output is only as good as the brief. And the magic happens when teams. Stop hoarding their best prompts and start sharing them.
SPEAKER_06We're live at OMR, and I am here with Director of Customer Success Central EMEA at Integral Ad Science, Maria Looks. How are you, Maria?
SPEAKER_04I'm good. Thank you, Nadia. Thank you for having me.
SPEAKER_06I'm super excited to talk to you a little bit about how you're using AI. But first, is there one important or interesting conversation that you've had so far at OMR?
SPEAKER_04As you can imagine, there have been plenty and there will be many more, as this is the number one topic of this event. I would say one thing that really stood out is talking about the responsibility that has to come with using AI in different directions. So this is something in combination with trusting your guts on what your human smart thoughts would give as an input when using AI, and that's the ideal combination.
SPEAKER_06Yes, and that's why we're here on the Still Human podcast, because we still need humans to do all of these processes and be the guiding light, let's say, behind them. So diving right in, um, what would be one of the most surprising things that AI has done for your team in the last six months, but not only business-wise, maybe also for your people?
SPEAKER_04So, what I could perfectly name as an example, what I didn't expect at all is using it for casual things like organizing a team event. So it makes repetitive tasks, but also things you would still put a lot of passion into uh very easily. So doing good things, um, motivational things, uh, meetings, parties, uh get-togethers become so much easier while overcoming barriers that would come from a language, nature. So language barriers are much more easy to overcome.
SPEAKER_06I remember we were talking about that, how if someone doesn't have English as a first language and they're wanting to write a certain way or approach maybe at a higher level, uh, you know, uh a type of conversation with a colleague, um, that is very helpful to have that kind of prompting and uh usage of AI tools that can help you. Indeed, it gives a lot of self-confidence as well.
SPEAKER_04Yeah. Because you can express yourself much better.
SPEAKER_06Yeah. So you also described the shift from Google search, which is the one word, two word keywords that we're putting into Google in the past now to searching in LLMs where you're almost writing a brief like you would be briefing a human. And so, what does that look like for you in practice? And why do so many people get it wrong?
SPEAKER_04In practice, it means I need to sit down and give some proper thoughts on what I would expect as an outcome, outcome before I even start prompting. So, in comparison to Google Search, where you would have played around, and we all have done that in the past for plenty of years, right? Uh just putting keywords in and see what comes out of it. Now you have to proper think if you want to get something useful out of it. Um, what do I actually want? And this is all ruled from creation and uh artificial Photoshop skills and everything uh you've been working on. Whenever you you expect a good outcome, it can only be as good as the brief.
SPEAKER_06Yeah. So if you put garbage in, garbage comes out, right?
SPEAKER_04As they say. Yes. And I do think um this is why many people do not get it right from the beginning. Uh, as you need to gather experience on that factor. You need to make yourself aware of the fact that uh artificial intelligence can make things easier, but still you need to know what you want and be precise in framing your expectations. Yeah.
SPEAKER_06So for people maybe who are still hesitating on using AI or using it more so in their work, what would you say to convince them to lean into the AI even more after your first coffee in the morning, when you think of the most boring and repetitive task you have, this is what you can overcome with AI.
SPEAKER_04Uh, it can make your life so much easier on things you do not want to do in a manual way anymore. Example. Um checking, uh in particular in our business, it's checking data every morning, see if there's anything alarming. Um, writing one and the same messages to different team members or clients. Um, and so that I think everyone can can find a perfect personal example for that.
SPEAKER_06Yeah, I love creating an agent that gathers all the ad tech media news. So I'm not going to all the different sites or going to all the different newsletters. That's been a game changer for me, and I can quickly scan. Awesome.
SPEAKER_04It's all perfectly set up after the first coffee in the morning.
SPEAKER_06Yeah, and that's what I do. I have my coffee and I'm like, let's see what's going on. Good stuff. Yeah, we'll take that one. Yeah, and it's pretty simple, it's pretty simple to build. Um, last question: how can leaders, teams, and companies expand their AI knowledge and the application at work so that it's both practical and safe?
SPEAKER_04I do strongly believe in sharing is caring. So at the moment, everyone is encouraged to play around and find your own best ways to make AI a useful tool. The next evolution stage is really sharing with your teammates, best practices, um, ready to make it a joint experience. This is super important.
SPEAKER_06And across different teams too, I would imagine, right? Because if someone's doing something in a completely different team, you don't know about it, how can you work together or do or learn from each other, right?
SPEAKER_04100%. And and that's the perfect example for how AI can bring communic human communication closer together again. Because if you exchange experiences about how using AI, it makes it more personal again. Yeah. Yeah.
SPEAKER_06Well, Maria, thank you so much for taking the time. And uh, let's stay human. Very good. Thank you very much. Thank you.
Rosa Markarian, Consultant for AdTech, AI, Sustainable Tech
SPEAKER_05Next, I spoke with Rosa Marcaryan, go-to-market leader for ad tech, AI, and sustainable tech. She brings the question almost nobody at OMR was talking about, or at least out loud anyway, and that's what's the energy cost of all this AI use and the data centers we're building? And why are we so excited about it?
SPEAKER_06We're live at OMR. I'm here with Rosa McCnarion, and I would love to know a little bit about what you're doing and why you're here at OMR.
SPEAKER_03Yeah, OMR, one of the most exciting conferences for sure, uh, in Germany. I'm here as an independent um consultant for advertising uh tech AI and sustainable uh tech companies to help them bring uh their exciting um products to Europe.
SPEAKER_06Amazing. Um, and yeah, we just were in the PayPal booth, which was really nice. And uh it's uh we were just talking sustainability is going to come back, even though it's gotten on the back burner, but it's such an important discussion. And what we're really here uh to kind of talk about is um yeah, apart from AI, we'll get into that. Uh what is something that from OMR you're gonna take away? A conversation that you were really excited to have here?
SPEAKER_03Well, there can't be a conversation at OMR without AI. Exactly. Um, so what uh one thing that really stood out for me, uh, we listened to a talk about how you can recognize if something is uh produced by uh AI or not, and that sparked a conversation about um are you hiding that you are using AI or are you proud about it? And um, my view on it, well, we should all be um proud about it and um use AI as like our partner in crime to achieve much more than we are able to achieve just um by ourselves. So when I hear um like statistics, oh, half of the content on LinkedIn is AI generated, yeah, sure, but the really good content is the one where AI helps you to generate it.
SPEAKER_06What uh what is your take then maybe on the creative side, if we have to label it, not label it, actual creative, the imagery or the video?
SPEAKER_03Yeah, um to be transparent about it. I'm a strong believer that um in every aspect of life, being um transparent and um honest about it um just gets you uh farther ahead.
SPEAKER_06Yeah. Well, love talking about this AI topic. Um there is still the human in the background though. So for you, what's something that AI can never replace?
SPEAKER_03Um it's the human expertise and experience of uh working for many um decades in a special field. So um take myself as an example. I um help companies to bring their um product offering from the US to Europe, so to localize it to the European advertising um industry. And I worked in this uh field uh for um big companies for many um years. And uh AI helps me to um localize the content, but then to review the content and to understand what needs to be adjusted, what resonates particularly well in Europe. That's something that you can only do when you have the relationships in the markets, when you know the um people, um the stakeholders that are important when you bring um people together, when you um bridge these um intercultural also differences that we have across um countries. And I do that currently for um PayPal ads, for example, and um I use AI a lot on a daily basis.
SPEAKER_06What's what's your favorite way to use it on your daily?
SPEAKER_03Um actually to um talk about um ideas. So whenever there is a new uh topic to understand um what are things that I um ideas or perspectives I haven't thought about. Using a sounding board as a sounding um board, exactly. Oh, and uh get many more presentations done than that's for sure.
SPEAKER_06Uh I know we're going off topic here, but I'm so interested because presentation is a pain point for me. It's just something particular that you use for your presentations that you're loving.
SPEAKER_03Yeah, totally, that saves a lot of time. Yeah, but yeah, then you also have to put on um like the lens of uh really do a good um check if um all the content, if it's just sounds nice, but uh if it's just uh yeah, MIP content pieces, or um if there's really something um valuable and deep behind it. So the sounding board is the most important thing.
SPEAKER_06Yeah. Uh and what is something, in your opinion, that is maybe missing from the AI conversations that were having?
SPEAKER_03Yeah, that comes uh back to my um other most important topic, which is uh sustainability, because I think uh we all talk about um AI and how it makes things way more um efficient, how it helps us, etc. But we uh don't talk enough about what are also the implications on the sustainability side behind it, like the um energy consumption that AI needs, because um we see in many studies at the rate that AI is accelerating uh right now, we don't have the energy to support that. Yeah, and uh that's something we also have to take into consideration and make um decisions when to use AI for um what? Do you always need to use the most sophisticated AI? No, sometimes it's just common sense, right?
SPEAKER_08Yeah, yeah.
SPEAKER_03Be smart about um how to use it from a sustainability perspective as well.
SPEAKER_06Is there do you see maybe more policy and regulation that's needed in that area as well, going into I mean, data centers are popping up everywhere. What's your hot take on that?
SPEAKER_03Um, I mean, I think the regulation is uh one part, but it will take a longer time until that's uh all um implemented. And in these areas, it always goes uh faster if there is um an understanding from the consumer point of view. And so that's why we need to raise the awareness for uh this topic as well when talking about AI.
SPEAKER_06That's that's a good point. Um, one that I didn't really think about, but it goes back to educating people how to use it, when to use it, and also our younger generation as well, um school systems and everything, educating teachers how to teach about it, how to use it instead of just banning it completely. Uh yeah.
SPEAKER_03Yeah, totally. So for the younger generations, that's um especially a big uh challenge because they also have to build up the um expertise um of how to also solve things without AI so that you can use AI as a companion and not completely just uh rely on as a new brain.
SPEAKER_06Yeah, excellent. Well, Rosa, thank you so much for the chat and uh have a wonderful second day here at OMR. Thank you so much. Thanks.
SPEAKER_05And
Michael Mörs, Founder & CEO SMA Development
SPEAKER_05to wrap us up, Michael Moores or Michael Moores, that's my German accent, founder and CEO of SMA Development and Podman, the tool processing around a hundred thousand podcast episodes a month. He's been deep in AI since 2017, and he's got some unfiltered things to say.
SPEAKER_06We're live here at OMR, and I am with Michael Moers. Yes. And I want to get you on record because we had such an interesting conversation the other night. Um, but first tell me a little bit what it is that you do uh and and what what is your title? What are you building?
SPEAKER_00Okay, so uh I'm Michael, founder, CEO of a company called SMA Development. Um basically we are building AI and data heavy products for our clients, different domains, different uh uh areas where our clients are from. And uh since I think 2021, uh we started uh our own product, uh, a tool called Podmon, which is um the first use case was Google Alerts for Podcasts. So you type in your name or your company, your brand, and you get an alert and uh a link directly into the podcast episode where people are talking about you right now.
SPEAKER_08Cool, yeah.
SPEAKER_00That was the basic one, and then after year after year, we're building up lots of analytics and lots of AI stuff around this transcribed podcast episodes, which are around about 100,000 episodes per month. Wow, and then you have a lot of metadata, lots of fusion, lots of stuff you can do on it, and that's basically really, really cool.
SPEAKER_06Yeah, and well, that's why I would love to pick your brain a second. Um, but before we get into the data and what you're doing with AI and audio and podcasting, which is my favorite thing. Um at OMR, it's it's been a wild and fun two days. Is there something that is the buzzword that you're hearing everywhere that you're just like, that's enough, I can't hear it anymore?
SPEAKER_00Um, so there's not one specific uh buzzword. I think a few so like beyond uh this and this. Yeah, we talk about beyond this and this. So this is like kind of okay, what do you want? Uh so when you look into the timetable, beyond everything, yeah, there's so deep insights which are not uh uh available there. And uh yeah, that's one thing, and the other is like uh every week, and also here at every booth we heard that something is over, and AI takes over all your jobs and all your business, and that's bullshit, and I know that. Yeah, and uh maybe hopefully someone will realize in the future.
SPEAKER_06Yeah, well that's that's a really important topic because we I it is a fact that a lot of um AI integrations and adoption companies have let there's companies that are letting a lot of people go, thousands of people go. Um, so it's but is that the right reason? Is the AI actually doing the job of those people? I remains to be seen. But uh yeah, it's it's that that's very true, and it's uh a hot topic that um we're gonna talk about on the Still Human podcast in the future. Yeah, I think that's such a dedicated like what is happening with the future of work and how we're using it. But more specifically, how are you using AI with all of this data and and especially in the audio field?
SPEAKER_00Uh let me just one take on this, uh, on this specific uh uh layoff stuff. Um so yeah, I think the story uh we're an innovative company and doing lots of AI, so we have to lay off people is much more interesting for stakeholders than we say, oh, we have to hire too many people and we have to lay them off now. So basically, I think this is a good opportunity for a good storytelling when you lay off people, but I don't think that is a causal uh relationship between these things. It's too early. We see how uh how complex adoption in AI uh for companies is. Yeah, and this short period of time, nobody has layoff people because of AI. I'm pretty sure it will come, no no matter uh uh it's not if it will come, it will come. Yeah, but I'm not sure right now all the layoffs you see. Uh there are different reasons for that. But it's AI stories better. Yeah, we are cool, hey. We're doing all before we have to do it. Yes, yeah. I think that's that's not uh only small truth.
SPEAKER_06Okay, interesting take, interesting take.
SPEAKER_00Yeah, I asked your question. So uh we are we were starting uh doing machine learning, predictive analytics, all this stuff uh around AI and large language models as well 2017. So basically, around about 10 years now, uh we built up a team, we built up infrastructure processes. So we before we were uh a data and software-driven business, they need software applications, and uh there was always something with AI or machine learning inside of the software stack, but I really, really uh uh fast realized that when you want to build AI, you need software, it's always together, and AI is just a little component inside of a bigger software stack, and this didn't change over the 10 years. Also with LLMs, uh, it didn't change. You need infrastructure, you need scalability, stability, quality management. That's not what uh that's what AI uh makes AI work inside of a company, and that's what we are doing. Yeah, for uh examples uh beside the case.
SPEAKER_06Yes, give me give me exact exactly examples. I want the details, this is the real time.
SPEAKER_00So uh we have different uh kind of uh uh clients. So for example, um Deutsche Bahn. Yeah, Deutsche Bahn. Uh we are doing several projects since uh 2018 with them. It was about uh predictions of punctuality, it was uh forecasting energy management, it was about um uh stabling uh uh trains uh on parking positions, and there was always some kind of AI inside of this, but it was many many processes and many digitalization stuff, and yeah, this is one example, another one is e-commerce, huge uh forecasting uh for stock uh uh inside of the uh um uh inside of the company. When do I get out of stock? When I do I have to purchase new goods, okay. Um, and another one is um for uh a big client for the telecommunic telecommunication sector, uh, is like um getting all the social media data together from all different sources and then uh make a recommendation system to say where should I put my media budget, which stores, which locations are right now very uh uh very good performing, which are not good performing, and where does it make sense to put my media budget in specific? And how to measure the outcome. So, was that good?
SPEAKER_06Uh yeah, how did that measure, how did that come out uh at the end if that was a good decision or not?
SPEAKER_00So that it's it's it's a really tough task because you link digital to offline yeah, worlds. You need to say, okay, which are the measurement points which I get from one store. Yeah, and you have uh frequency data, how much how many people going inside and outside of the store? Uh you have um uh uh transactions of the store as well, and you have uh reviews, uh Google reviews.
SPEAKER_06Yeah, who's saying what?
SPEAKER_00Quality measurement, yeah, quality indicators. Uh um, and we find out uh when we see it's like time series analytics. When we see okay, we have different uh output signals or input signals, maybe ad campaigns or communication campaigns or store activation stuff, so promo people. Can we see uh statistical uh differences in in the behavior of the normal time series? Okay, and you can make lots of mathematics behind it, and then you can say, Okay, this was powerful for this, or the performance was good on this store, but not on this store. Next learning circle, take this store with this. Uh um health uh uh KPIs because your meetup actually will make the most out of it in this area and not in this area and stuff like that.
SPEAKER_06Cool. So in that, is there something in particular that now with the advancements we're seeing in AI, or also we could go down the adjunctive commerce route, but that that's maybe another topic for another time. But um what what's really changed then um in the AI application that you're seeing and how you're using this data and collecting this data and and and then making reporting and measurable intelligence decisions with it. Is there something really that's changed in the last six months?
SPEAKER_00Uh there's so much change. Yeah. So even for us, and we are also doing lots of research uh projects, so we are partnering with lots of big research institutes, finding out uh where are the boundaries, where are things going very well, but where are the obstacles in new models or in new setups, new technology stack tags. But uh even for us, it is really heavy to get uh before before the wave or serve the wave because there's so much new stuff all day long, and uh yeah, you you should stop working to get on to understand what all the new stuff is. And um, I think you have to have a look at from different perspectives. So for a private person, there's so many opportunities with AI in their private life, yeah, crazy stuff. And this is the biggest change uh open AI did 2022 uh because there were models before, but they didn't invent them so they've been but they just they weren't in that format seven years on these models and release lots of models, and then they decide per accident okay, let's put it into the center of the customer. So let's put it to a D2C uh audience before it was data science, very academic, very uh nitty-gritty coding and stuff like that. The tech and the nerdy people and all that's only naughty stuff, and now everybody has can interest has these capabilities, they can build apps, they can build websites, they give like that's so a lot of possibility, yeah. And it also changes our uh word uh every day. So we are doing AI coding as well. Uh but we have senior developers doing this, they can understand what the AI is doing, and this is a really, really big part. Because um I think you should not trust uh AI too much because in every task, design, coding, communication, it's uh hallucinating as well. Yeah, it's it's in the models itself, and when you don't know how what code looks like and what is good and what is bad, you pump out so many applications, there's so much stuff that you have lots of vulnerable rebellious uh in there, uh lots of critical things. And uh yeah, maybe this is a really security issue as well, then.
SPEAKER_06Yeah, you have to be very careful. Maybe you can build that website in two hours, yeah. But it's always a good idea, like you said, to have the idea to think, okay, should I bring someone in maybe um from a UX perspective? Uh is this working? Is this vulnerable uh to malicious hardware or software, whatever? Um uh from an engineering point of view, because you put in a maybe a maybe even a detailed prompt, but uh that doesn't necessarily mean, like you say, that's gonna be a completely 100% safe and ready to go website.
SPEAKER_00And and uh everything looks really good from these models. They are designed to do you a favor.
SPEAKER_09Yeah, yeah.
SPEAKER_00And and uh uh they are really good in uh explaining to you why they are really good in this, yeah, but you should not trust it too much because um yeah.
SPEAKER_06The human and loop in this, yeah, human loop is this, you just have to you still have to think about what that means um from a uh a safety point of view, security point of view, yeah, um a design point of view. Yeah uh so yeah, we still need humans.
SPEAKER_00Yes, still still very, very uh uh important. Uh even and uh even with my team, so we are 22 people, even with my team, I see how people rely, start to rely and trust them, and then you see how many uh errors or mistakes happen. And we see okay, we have to be really open-eyed and really uh uh concentrated on the point. Yeah. Because when you let the agents run, they will do lots of stuff and burn lots of tokens, and then you have lots of work afterwards to clean up the mess.
SPEAKER_06Well, we're seeing that is and that's uh an interesting point because and also what I've been reading as well is um you sometimes, especially in going back to like how it's being used in within teams and in an enterprise or a working situation, is um it maybe will produce so much. It's like, okay, great, I did I got this like 25 page more, but like that's like so much. Like I didn't need that much, and now you have the the humans that have to like kind of go through and decipher it and and take out actually what's not needed because it was just like vomiting AI, everything onto you.
SPEAKER_00And when you're looking into enterprise processes, yeah, uh, how many percent of one project is coding actually? Yeah, maybe 50%. The rest is communication and uh stakeholder management and uh consulting and understanding the problem and just get insight of this huge uh corporate stuff, and uh that's still a task uh which is really important. And uh coding is not more the most important stuff because yeah, it's commodity, uh, but doing the right thing, coding the right thing, which fits into the processes and into the minds and it gets trust from the people inside of the corporate, that's the most important task.
SPEAKER_06Yeah, perfect. Michael, thank you so much. This has been an incredible conversation. Thank you for sharing all your amazing insights and uh have a great rest of your OMR.
SPEAKER_00You too. Thanks for having me.
SPEAKER_06Thanks.
SPEAKER_05That's a wrap on today's episode. Thank you for tuning in. If this conversation made you think or made you want to send it to someone who needs to hear it, please share it. That's genuinely how this show will continue to grow. And if you have an extra two minutes, please leave us a review or comment. It really makes a difference to who finds the Still Human podcast. I'm always looking for the next great conversation. So if there's someone you think I should be talking to, let me know. Drop it in the comments, send me a DM on LinkedIn, all the info is in the show notes. And if we're not connected yet on LinkedIn, let's chat. I'm there most days sharing what I'm musing about and working through. And that's usually where the conversation continues between episodes. Hit that follow or subscribe button to stay up to date with all the amazing conversations happening here on Still Human. New episodes drop every other week, and I don't want you to miss what's coming up next. A big thank you to my sound engineer Morgan Ball for making me and this podcast sound brilliant. Always, even on five hours of sleep. Thanks again for listening and giving a damn about this stuff. Until next time, keep feeling, performing, doing, speaking out, and above all, questioning. Because after all, we are still here.