So I’m sitting here with my third bourbon of the morning—don’t judge, it’s research—reading this interview with two German AI consultants, and I’ve got to say: these guys actually get it.
Timm and Johannes from some outfit called “disruptive” in Munich just dropped the most honest line I’ve heard about AI implementation in months: “KI ist kein Tool-Thema, sondern ein Kulturthema.” That’s German for “AI isn’t a tool problem, it’s a culture problem,” and brother, that’s the whole goddamn ballgame right there.
See, while everyone’s been huffing the hopium about how ChatGPT’s gonna save their business, these two have been watching companies light money on fire. My favorite story from their interview? Some construction company bought 300 Microsoft Copilot licenses at the start of 2024. Cost them 100,000 euros a year. Six months later, exactly two people were using it.
Two. Out of three hundred.
That’s not a technology failure, folks. That’s a human failure wrapped in a corporate purchase order. And it’s beautiful in its stupidity.
The Four Horsemen of AI Implementation
These Munich boys break down AI strategy into four pillars: Technology, Training, Governance, and Mindset. Most companies nail maybe one of these. Usually technology, because that’s the easy part—you just throw money at vendors and feel productive.
But here’s where it gets interesting. Johannes talks about companies that build their own internal GPTs for data privacy reasons, which sounds smart until you realize they’re basically building a worse version of something that already exists, then wondering why nobody wants to use their janky knockoff.
It’s like brewing your own whiskey in the bathtub because you don’t trust the distillery. Sure, you might end up with something drinkable, but more likely you’ll go blind.
The real kicker comes when they mention legacy systems. Timm says they’ve walked into companies as AI consultants and walked out as ERP consultants. That’s like showing up to treat someone’s headache and discovering they’ve got a brain tumor. The AI stuff was never the problem—the problem was that their entire digital infrastructure was held together with duct tape and prayers.
The Prompting Prophet
Now, Johannes claims he’s not the “master of prompting” despite apparently being exactly that. He’s got that European humility thing going on, which is refreshing compared to the American tech scene where everyone’s a “thought leader” after writing three tweets.
His take on prompting strategy is actually useful: be precise, don’t overload the model, work iteratively, stay in dialogue. It’s not mystical, it’s just practice. Like learning to drink whiskey—you start with the cheap stuff, work your way up, and eventually you can tell the difference between a good pour and rotgut.
But the real wisdom here is when he says AI isn’t a replacement for professionals. It’s a sparring partner. That’s the line that separates people who understand this technology from people who think it’s magic.
I’ve been using AI as a sparring partner myself—mostly to argue with when I’m too drunk to find a real human who’ll put up with me. The machine never tells me to go home, which is either a feature or a bug depending on your perspective.
The Refrigerator Moment
Johannes tells this story about standing in front of his fridge, not knowing what to cook, so he takes a photo and asks AI what he can make. First result was garbage, but after some back-and-forth, he got a decent recipe.
This is the future, people. Not self-driving cars or robot butlers—it’s AI telling you how to use your leftovers. And honestly? That’s more useful than 90% of the sci-fi predictions we’ve been force-fed for decades.
Timm makes a good point here: in two years, asking if someone uses AI in their personal life will sound as dumb as asking if they use the internet or a mobile phone. He’s right. We’re living through one of those transition moments where the weird becomes normal, and nobody can quite remember when it happened.
The Corporate Comedy Hour
The interview reveals this beautiful pattern: companies with “high AI maturity” are hiring way fewer entry-level people. Which makes sense—if AI handles all the grunt work, you don’t need grunts. But then Timm tells students that people who really understand AI will always find jobs.
So which is it? Fewer jobs or guaranteed employment?
Both, probably. The middle’s getting hollowed out. You either understand this stuff well enough to be valuable, or you’re competing with a $20/month subscription. There’s no in-between anymore, and that’s terrifying if you’re trying to break into the industry doing the same copy-paste work that kept people employed for decades.
The Vision Thing
When they start talking about the future, things get interesting. Johannes sees better internal systems, AI understanding corporate design, realistic avatars, and AI-generated music flooding Spotify. Timm’s betting on AI-powered audio and vision interfaces—smart glasses, headphones with live translation, contact lenses with displays.
The teleprompter example hits hard. Why would you need one when you can have your speech displayed directly in your field of vision? Another product category about to get murdered by progress.
But here’s what neither of them says explicitly: all this convenience comes with a price. When AI can generate convincing music, video, and text, how do we know what’s real? When everyone’s wearing smart glasses feeding them information, what happens to actual human interaction?
They touch on it—Johannes mentions the challenge of distinguishing real from generated content—but they don’t dwell on it. Can’t blame them. Nobody wants to be the buzzkill at the AI party.
The German Efficiency
What strikes me about this whole interview is how practical it is. No grandiose promises about AGI or the singularity. No panic about AI taking all our jobs. Just two guys who’ve been in the trenches, watching companies struggle with change management disguised as technology implementation.
They even built an online learning platform—the “disruptive KI Akademie”—because they realized most companies can’t afford to send everyone to multi-day workshops. That’s smart. That’s understanding your market. That’s also kind of depressing, because it means even the basics of AI literacy are now something you need to pay for and schedule time to learn.
But at least they’re honest about it. They’re not selling snake oil. They’re selling training wheels, and they’re upfront about the fact that you need them.
The Reality Check
Their upcoming keynote is called “Myths, Misunderstandings, and Real Progress: Generative AI in the Reality Check.” I’d actually go to that if I weren’t allergic to conferences and Nuremberg.
Because that’s what we need more of—reality checks. Less hype, more honesty. Less “AI will change everything” and more “AI will change some things if you actually do the work.”
The construction company with 298 unused licenses is the perfect metaphor for where we are right now. Everyone’s bought tickets to the future, but hardly anyone’s actually boarding the train. They’re standing on the platform, ticket in hand, wondering why nothing’s happening.
Meanwhile, the people who understand that this is a culture problem—not a tool problem—are actually getting somewhere. They’re the ones asking hard questions about training, governance, and mindset. They’re the ones willing to admit that sometimes you need to fix your basic infrastructure before you can bolt AI onto it.
The Bourbon Truth
Look, I’m not saying these German consultants have all the answers. But they’re asking better questions than most of the AI evangelists I’ve encountered. They’re not selling salvation, they’re selling pragmatism. And in a field drowning in bullshit, that’s worth something.
The truth is, AI is already here and it’s already changing how we work. But it’s not magic, and it’s not going to fix your broken processes. If anything, it’ll expose them faster than you can say “digital transformation.”
So yeah, maybe the Germans are onto something. Maybe AI really is a culture problem. Maybe the technology is the easy part, and the hard part is convincing humans to actually change how they do things.
Or maybe I’m just drunk and overthinking an interview about a conference in Nuremberg.
Either way, I need another drink.
—Henry
Wasted Wetware: Tomorrow’s tech news, today’s hangover
Source: "KI ist kein Tool-Thema, sondern ein Kulturthema.": KI DOAG e. V.