Back to the Future of AI – a podcast journey about artificial intelligence in companies and society with AI pioneer Alexander Tam. Today we welcome a guest who shows that technology and entrepreneurship are more than just business; they drive innovation and progress. Our guest is co-founder of Flix and as Chief Information Officer and Chief HR Officer is responsible for technology and HR. His heart beats for education, innovation, and entrepreneurship. As an active investor and advisory board member, he supports companies like Baby One and Unity. He also shares his knowledge as a lecturer at Friedrich Alexander University Erlangen-Nuremberg and advises the MCI Management Center Innsbruck, the Ansbach University of Applied Sciences, and the University of Hamburg. As a co-shareholder of Startup TEAM, Germany's largest NGO for entrepreneurial education, he promotes the founders of tomorrow, and at the Code University he brings his passion for digital education as a trustee. Welcome to Back to the Future of AI, Daniel Krauss.
Yes, dear Alex, I'm glad to be here, I'm really excited, and thank you for taking the time. Anytime. That's not true, we had to reschedule a bit, but always happy to. Very nice.
You might know it if you've seen a few episodes – we travel through the past into the future. I remember our panel we had together; how long ago was that? Two years or so. I think a lot of what we experience today we already discussed on stage back then. But before we dive into that, tell me how you got into IT, data, AI – how did it start for you?
It actually started quite typically for our generation. My parents ran a construction company, and in the registry there was an old monochrome computer for writing invoices. I abused it to print funny graphics instead. Then came 286, 386, 486... In fifth grade, against my parents' will, I got my first real PC – a black Windows 95 machine from Escom. My uncle sat down with me and that's when my enthusiasm turned into something productive. We had LAN parties that were a science in themselves with terminators and TCP/IP. I studied business informatics and realized my sweet spot was the interface between technology and people – translating requirements and understanding architecture. I was never a great programmer; they joked I could only solve things with if-loops. But I could make sense of things. That's why at the beginning of Flix, although there was a lot of tinkering, I found great colleagues early on. Flix is about 12 years old now. Now with responsibility for technology and HR, it's my perfect fit.
And what did you play, if I may ask?
We started with Quake 2 and Red Alert, then Counterstrike and Starcraft. That fascinated me. Then I started at Siemens and got stuck. Over time, especially during my studies, I realized the interface between technology and people was where I felt comfortable. I was never a good programmer, but I could structure things. That's why at Flix, we had a lot of tinkering at first, but I found great colleagues. Flix is about 12 years old now.
Very exciting. I flinched at the games. I think at least 70-80% of people have such a past. I remember dealing with memory management on our 386 because nothing worked. You went into autoexec.bat and had no internet to ask. You found out which dad knew computers. My uncle was a super nerd. That playful access is important – we used to screw computers together and solder just to get the optimum. That playful learning is so important because it's easy access. I'm not happy with our current education system – it's too rigid. Back then you got deep into the basics, close to assembler. I hope that doesn't get lost today when everyone just installs an app. For AI, it would be good if humanity understands how it works and doesn't just consume it. With my kids, I make sure screen time includes something productive, not just consumption.
That's exactly it. Playful learning is crucial. My first mobile was a Sony C1 – it had a terrible UI but no antenna. I always liked being non-conformist. I had a Windows Phone even when it was dead. I later switched to iPhone because of the ecosystem, but now Pixel phones are also good. The Blackberry story is fascinating – I saw the documentary on Netflix. My first business was an internet café in our village. That got me hooked on the next hype.
Exactly. The combination with the internet was huge. My first business was also an internet café. Those were wild times. That curiosity led us to data-driven thinking at Flix from day one.
At Flix, we were never arrogant that we knew everything. We saw the market open up after deregulation and said, 'Okay, we don't know exactly how this works – we have no domain knowledge of transport – but we have business and technology know-how.' We started collecting data from day one. We built crawlers and used Excel to figure out where people wanted to go. We were data-driven because we didn't know better. Our first hypotheses, like Nuremberg-Regensburg, failed completely. But we learned that data is essential. The algorithm and models are important, but without meaningful data, you're nowhere. That's the foundation – we can rightfully say we are a tech company, not a bus company.
That's so cool. I can listen to you talk about this. It's like someone saying the earth is not a flat – we've figured that out. But even after ChatGPT, there are still people who don't get it. Why is that? It's because it's not part of their lived reality. Look at the recent election – record turnout, but also many first-time voters moving to extremes. That's about lived reality. You and I, we had our first websites back in the day, party events in Nuremberg. That was before the internet was mainstream. It only became part of life between Yahoo and Google. Before that, it was just us nerds. At Flix, we were praised for bringing internet to buses in 2013. And I said, 'Nice, but it's just attaching some boxes.' Now it's standard. So it takes time for LLMs and agents to really enter lived reality – and that's okay.
I agree. The deepsieg moment is interesting. It's good that tools are becoming cheaper. I use both ChatGPT and Google, because some questions make no sense on GPT. You need to connect things properly and not just throw the dumbest question at an LLM to produce costs (and CO2). The path is paved – there's amazing stuff every week. When someone asks where the company will be in ten years, I say, 'Who talks about ten years? We can talk about ten minutes or ten days.' It's crazy.
I also find it fascinating how political AI has become. In the past, we just tinkered. If you told people you worked in internet, they thought 'internet stuff.' Now it's highly politicized. Our company name was almost 'Advanced Business Analytics Consulting' – that sounds like the 90s. But we dodged that bullet.
