#workculture24 | Übernimmt jetzt die KI? Opening mit Birgit Bohle

Yes, thank you very much, dear Linda, for the kind welcome. I'm really happy to be here; I was at Work in Culture before, but virtually, and now it's wonderful that despite various strikes, so many people have made their way here. A warm welcome from me, nice that you're all here. Also a warm welcome to all the HR professionals I suspect are here. I'm excited to speak today about a topic that is on everyone's lips: artificial intelligence, which is extremely polarizing. There are the apocalypticists who say AI will bring about the end of the world, and others who say it's the only salvation for productivity and economic prosperity. There are people in society and in companies who fear AI, and business leaders who are investing heavily in AI. What is certain is that AI is here to stay, and it's also clear that we still have no idea where AI will actually lead us. I'm sure we'll talk about it a lot from this stage, not just this year. What I'd like to discuss with you today are five theses – five theses on how we at Deutsche Telekom look at AI, five theses about the impact of AI on the world of work. Let me start with the first thesis: as Deutsche Telekom, we want to become the leading digital telco, and for us, AI is an absolute must. We see it as a huge opportunity, and I believe it's also a huge opportunity for Germany. I believe AI is the innovation and productivity boost we really need. We are a major industrial nation, but when we look at all the performance indicators – and I won't launch into a lament about Germany – we see that things aren't running smoothly: economic growth, companies leaving, a shortage of skilled workers. So we need productivity improvements, and that means we need innovation, and that is AI. I also believe it's not too late. We had Jonas Andrulys from Aleph Alpha here at transform: 'Made in Germany AI, made in Germany.' And my highly esteemed colleague Ferry Abolhassan – that's a spoiler for you later on the transformation stage – the head of T-Systems will talk about AI. So we have companies that can drive this forward, and researchers say that developed economies like Germany can massively benefit from AI, with a 35% productivity increase forecast by Accenture for 2035. When I look at us at Deutsche Telekom, we use AI in very different fields. Take fiber optics: we are currently building 2.5 million fiber optic households per year. We couldn't even plan that without AI. For that, we have vehicles called T-Cars equipped with cameras that film the surroundings and calculate the optimal routes using AI. The result is more efficient, and hopefully you'll all get a faster fiber optic connection from us soon, if you don't already. People used to do this work manually; their planning activities have changed and are now supported. That brings me to my second thesis: AI transforms and revolutionizes individual work steps and processes radically. There are various studies and figures on how many jobs will be lost or created by AI, and honestly, no one really knows. What we do know is that many processes will change, tasks will change, some jobs will become less common or disappear, and new ones will emerge. For example, I didn't know the term 'prompt engineer' ten years ago – forgive me – but today I know what it is. Currently, AI is a tool that changes and accelerates individual tasks. Take Microsoft Copilot, which I've been using for a few months. When it comes to quickly summarizing a document, it's one click and there's a suggestion. Then I improve that suggestion in maybe 20 seconds, saving me 10 or 20 minutes of work. Steve Jobs once said a computer is a bicycle for the mind. If we transfer that to AI, I'd say AI is a racing bike in a wind tunnel – it helps us go much, much faster with our own brains. But we still have to learn to ride the bike, we still have to keep our balance. And we invest heavily in that at Telekom. A colleague just said we've equipped 40,000 people with AI skills – by now it's 80,000. So we're really trying to give many, many employees the basic tools and understanding of AI. Looking at my own area of responsibility – as an HR person, a lot is about learning – we are currently introducing a talent intelligence platform called 'fold'. We're piloting it in five countries, from Austria to India. Based on AI, a talent marketplace is being developed where skills and profiles are automatically matched with projects and positions within the company, creating a marketplace. At the same time, employees receive personalized learning offers through AI so they know what they still need to learn for a particular job or project. That is AI in concrete application at our company. We of course also expect our leaders to be fit in AI. That brings me to the third thesis: AI will not lead the world – that remains for us humans. I just talked about how work becomes more efficient. According to one study, 70% of employees and leaders hope that this will be the case, that they will be relieved. That's great; as leaders we then have more time for real leadership. But what is that exactly? What can we do better as human leaders than AI? You might quickly say empathy – that always comes up. I tried it: with a coaching AI named Amy. I asked Amy how I should talk to a difficult employee. I was quite surprised because Amy was far more empathetic than I expected. She asked me a lot of smart questions and gave me some very helpful tips for the employee conversation. I'd say if everyone followed those tips, many people would be a bit better at the basics of employee conversations than they perhaps are today. So an AI coach – but who actually wants that? Market research here really surprised me; it's from Oracle. People were surveyed internationally about whether they would rather turn to a human or a digital leader with psychological problems. The answer: 82% prefer the digital leader. Over 80% also assume that a digital leader would judge them more objectively than their own. I think that should make us a bit thoughtful. Because what else does that say than that AI – a digital leader – enjoys more trust in some areas, is perceived as more objective? That can't be right. So AI challenges us as leaders to become better. What I see as the real task of leadership: as leaders, we set the direction, we are courageous, we make decisions, we are truly empathetic – not just pretending. And in the end, what AI will never take away from us: we take responsibility, we take responsibility for our decisions. And I want to talk about responsibility in my fourth thesis: it's up to us to shape AI responsibly as a tool. We need a space and a framework to design AI responsibly – that applies to regulation and also to us in companies. What did we do at Telekom? Very early, in 2018, we developed guidelines for the ethical use of AI. They include things like transparency: our customers should always know when they are dealing with AI. We agreed on that in 2018. We developed an AI manifesto with the works council. That manifesto states that we are open to AI, but please, always let a human make the final decision on a personnel matter or staffing decision. We need this framework, and we also need it regulatorily, because we all know the risks of AI. We're in an election year; we know about deepfakes, how social bots can influence elections, and the biases of AI. Recently, we held an AI conference in Athens, gathering all the experts from the company. I tried prompting, trying to create an image of the Deutsche Telekom AI conference in Athens. What happened? No women in the picture. I said, 'please with women.' Okay, what happened? No more Magenta in the picture. I said, 'please with Magenta.' What happened? No more women. Is that now the problem of AI? Not really; it's the problem of our real world. Just look around the room here. AI reflects what reality is today. So it's both a risk – to reinforce that – and an opportunity – by training with different data, with unbiased data, to overcome risks. I come to my last and most important thesis: AI can enrich and improve our lives. I mentioned at the beginning what opportunities AI brings for the economy, for Germany. But for that, we really have to see it as an opportunity, and technology as an opportunity. And enthusiasm for technology and innovation is not always in the cradle for us Germans. If we look at the Edelman Trust Barometer for 2024, which researches trust, it says 50% of Germans reject AI – 50%. My board colleague Srini would say 'German Angst.' And we know that fear is in our genes; it's a good survival strategy from the Stone Age, when it was helpful. But AI is not a saber-toothed tiger. We shouldn't run away from it. Rather, AI is a tool for efficiency, innovation, and a better life – just like the spearhead or the hand axe in the Stone Age were tools for a better life. That's how we have to understand AI. We need courage for this technology, and the prerequisite is trust. People must trust AI. As companies, we have a very important task: it starts with demystifying AI and showing that it's already part of everyday life in many areas – be it face screening on smartphones or route planning. But we also have to show that AI not only makes our lives more efficient and faster, but truly better. I've already talked about the fiber optic connections that hopefully make life better. A great example from T-Systems is a project we recently awarded. It involves an AI solution we developed for a hospital, using health data to improve the condition and prognosis of people in intensive care. That, we hope – we're in the pilot phase – will save lives. That's how AI really makes our lives better. We have to empower our employees – I've talked about that. We have to establish rules for responsible use. And above all, we have to promote curiosity and enthusiasm for AI again and again. We have a guiding principle at Deutsche Telekom: 'stay curious and grow.' We expect that from all our over 200,000 employees. Then we can expand our horizons, then we can really see AI as an opportunity. Because in the end, one thing holds: we want to make AI with people, for people, so that our lives become better through it. Thank you very much.