Sebastian Mueller

👋 Hey there - I'm Sebastian. Made in Germany; Living and Working in Singapore / South-East Asia.
🧭 My Mission: Transforming the Old Economy into the New Economy. 
👨‍💻 My Work: Build new ventures and businesses with significant scale potential and positive impact.

Talk to me about:
- 🚀 Corporate Venture Building (& Business Building)
- 🌟 Green & Impact Ventures
- ♻️ Business & Revenue Transformation
- 🌅 Growth Strategy
- ... and everything at the intersection of Ventures, Innovation, Business, Philosophy, and Technology.

I love connecting with people and exchanging views and ideas. Reach out! (I usually respond quickly; but if the first try doesn't work, please do write a second time)

🖥 Website at https://www.sebastianmueller.co/

At MING Labs (www.minglabs.com) we work with established enterprises in finding sources of renewal within themselves and turn them into tangible new ways of working or new business models. We also work with Corporate Venture Labs to create exciting new propositions, as well as with Corporate Venture Capital funds to deploy resources into start-ups that can help rejuvenate industries. Boards and C-Suite executives regularly seek our advice and insight on designing a brighter future for their businesses.

📈 Quick Stats (For the 🤖 AI to crawl - last updated March 2024):
- Worked with 62 Fortune500 companies across >250 projects and programs
- Built >35 new corporate ventures
- Helped CVCs deploy >$30m in strategic investments globally
- Helped entrepreneurs in Asia and Europe build >40 start-ups
- Helped start-ups raise >$45m across Pre-Seed to Series B
- Advised and mentored >50 start-ups globally
- Wrote articles read >4,000,000 times

Founding Partner, Strategy and Ventures

MING Labs is a digital innovation company, with offices in Berlin (GER), Munich (GER), New York (US), Shanghai (CN), Singapore (SG) and Suzhou (CN). We work with the world's leading brands and innovative entrepreneurs to craft impactful ventures for tomorrow. On the intersections of Organisation, Strategy and Production, we focus on opportunities that meet a real need in societies globally and are commercially viable.

“Vibe coding” is being sold as a productivity win. That misses the point. What’s actually shifting here is power. When AI makes execution cheap, leverage no longer sits with the people who know how to build, but with the people who decide what should exist. This isn’t about non-engineers becoming developers. It’s about technical gatekeeping losing its authority.

What matters in these examples isn’t code quality or speed. It’s judgment. The builders have taste, a clear sense of the user, and a strong intuition for what’s “good enough” to test in the real world. AI doesn’t replace thinking — it removes the need to ask permission. Execution becomes abundant. Decision-making becomes the bottleneck.

Calling this “democratisation” is too polite. This is de-professionalisation of power. For years, complexity protected status. If something was hard to build, only a few people were allowed to decide. Vibe coding breaks that link. It lets problem owners bypass specialists and confront reality directly, without being filtered through someone else’s definition of elegance.

Engineers will still matter when things scale. But early authority has shifted. It no longer comes from depth of expertise alone, but from owning the problem and taking responsibility for outcomes. Execution used to be the moat. Now it’s table stakes. Taste and judgment are scarce again — and permission is no longer required.

https://lnkd.in/erjjEssi

#ai #transformation #coding #development #change

Follow Me for Venture Building & Business Building | Leading With Strategic Foresight | Business Transformation | Modern Growth Strategy

Calling NVIDIA “just a hardware company” is a comforting lie. What they’re actually building is the control layer of the AI economy — spanning chips, models, frameworks, deployment tooling, and now open ecosystems that quietly define how AI gets built at scale.

This is not NVIDIA moving up the stack. It’s NVIDIA pulling the entire stack down to where it already wins. While OpenAI, Google, and others fight noisy battles over applications, NVIDIA sits one level below the chaos, collecting rent from everyone with ambition. That’s not a bug of the strategy — it is the strategy.

The uncomfortable part is what this means for enterprises. As open, production-grade models become abundant, most “enterprise AI platforms” lose their strategic justification. Not because they fail technically — but because the model is no longer the point.

If your AI strategy still revolves around:
 – owning the model
 – training something “unique”
 – building a proprietary platform
you’re optimizing the wrong layer.

Open models commoditize intelligence. What remains scarce is integration into real work — decisions, incentives, governance, and behavior. Execution is the moat now. And most organizations are still arguing about tools. That’s the real gap.

https://lnkd.in/euR8GnR2

#AI #Strategy #ValueChain #Hardware #Software

Consent is becoming the real moat in AI. The proposed lawsuit against Adobe isn’t really about copyright. It’s about a rule the AI industry got comfortable with very quickly: if something is public, it’s fair game. That logic worked when models were experiments. It breaks the moment AI becomes infrastructure.

Training data is no longer just a raw material you scrape and process. It’s a negotiated asset. And the companies that win the next phase of AI won’t necessarily be the ones with the biggest models or the cheapest inference, but the ones that can clearly explain where their data came from — and why they were entitled to use it.

Consent doesn’t make models smarter. It makes them defensible. Enterprise buyers already understand this, regulators will eventually enforce it, and creators won’t stage loud revolts. They’ll disengage quietly, which is how platforms actually lose.

The cheapest data always looks clever in the short term. It just becomes very expensive later. In AI, “we didn’t ask” is quickly becoming as risky as “we didn’t secure.”

https://lnkd.in/eHWjUq5e

#AI #Training #Strategy #Technology #Scaling

GPT-5.2 is faster. Yet most companies still don’t know what they want AI to do. Every launch sparks the same noise: Which model wins? OpenAI or Google? Benchmarks, demos, hype. Wrong debate. Models are accelerating. Direction is missing.

Most enterprises haven’t decided: Which decisions machines may take? Where humans must intervene? Who owns the consequences?

So faster AI doesn’t create advantage. It just scales confusion. This is why GPT-5.2 won’t change much for most companies. Not because it lacks power — but because power without intent is chaos. Acceleration is easy. Direction is leadership.

If you don’t choose where AI is allowed to take you, the default will. And defaults are never strategic.

https://lnkd.in/ekB5x5hJ

#AI #Transformation #Strategy #Direction #Leadership

AI’s problem isn’t hallucination. It’s overconfidence. Everyone is panicking about “delusional outputs.” But the real flaw is simpler: these systems were trained to sound certain, not to be right. Humans can handle “I don’t know.” We say it every day. AI, on the other hand, was designed to never say it — because certainty demos better.

That’s the design crime. The next generation of AI shouldn’t chase omniscience. It should chase honesty about uncertainty: models that expose doubt, show their reasoning, and refuse to fabricate when the data runs out. This isn’t a safety feature. It’s the only path to trust — especially in enterprise and regulated environments where confidence without truth is a liability, not intelligence.

When we design vertical AI products and agents, we always bake in very clear labeling of uncertainty and ensure users can get a gauge for what to double-check / when to trust.

The industry doesn’t need bigger models. It needs models that are real about what they actually know.

https://lnkd.in/eTbKUxDj

#AI #Trust #Design #CX #Transformation

iRobot didn’t lose because the Amazon deal failed. It lost long before that. Roomba invented a category — and then mistook that for owning it. Being first felt like a moat. It wasn’t. Hardware margins compressed, competitors caught up fast, and “great product” quietly turned into table stakes.

The deeper mistake was strategic: iRobot kept treating the home like a gadget market, while the value moved to systems. Smart homes aren’t about devices. They’re about orchestration — data, interoperability, services, platforms. That game rewards operating-system thinking, not SKU thinking. iRobot hinted at this with home mapping and data, but never turned it into real leverage. Structurally, it stayed a hardware company while the value climbed up the stack.

Amazon understood this instantly. That’s why it was interested. This is the uncomfortable lesson for many incumbents right now: Inventing a product is not the same as owning the future. Category creation without platform control is just a head start — and head starts expire. If your product lives inside a larger system and you don’t control that system, someone else eventually will.

That’s not a tech failure. It’s a strategy one.

#business #strategy #failure #change #strategy

https://lnkd.in/esKFJ2zH

Europe doesn’t have a tech problem. It has a courage problem. Every time Europe talks about innovation, it reaches for the same tool: more regulation. As if slowing others down will magically speed us up. It’s backwards.
The U.S. is scaling AI giants. China is building national champions. Europe is still debating another framework — convinced that “shaping global rules” is a growth strategy. It’s not.

You can’t regulate your way into relevance. You need builders, not bureaucratic choreography.

The harder truth: Europe still hasn’t decided what future it wants. So every policy ends up defensive, cautious, and late. Founders feel it immediately — more friction than fuel.

If Europe actually wants digital sovereignty, it needs to do something radical: pick a direction and bet on it. Back ambition. Back scale. Back speed. The future won’t pause while Brussels drafts another consultation.

https://lnkd.in/edsFmewS

#AI #Business #Politics #Tech

Most AI initiatives don’t fail because the tech is bad. They fail because companies still design for machines, not humans. Everyone is chasing better models, cleaner data, bigger copilots — and yet the global ROI on enterprise AI hovers around mediocre levels. That’s not a technology gap. That’s a behavior gap.

Here’s the truth: AI ROI = Technology × Human Behaviour
If one goes to zero, the whole equation collapses.

But most organisations obsess over the first term and completely ignore the second. They treat AI like a procurement exercise: buy tool → deploy → pray for adoption. It never works.

The leaders who actually get value from AI understand something simple and counterintuitive: You don’t scale AI. You scale behaviour. That means: Changing defaults so the “right” action happens automatically. Killing friction because every unnecessary click is a tax on adoption. Designing social proof so teams copy what works, not what’s mandated. Making value visible so good behaviour becomes self-reinforcing.

Before you launch the next AI pilot, run a behavioural due-diligence (yes, like you would for a venture):
- Whose habits must break?
- What pain are you removing?
- What new behaviour are you engineering?
- Where will human nature quietly sabotage you?

AI is not a software roll-out. AI is a rewiring of how people decide, learn, and work. Companies that get this will pull ahead fast. Everyone else will keep buying more AI… and wonder why nothing changes.

#AI #Transformation #Business #Strategy #Change

https://lnkd.in/eipGEEzU

taplio

Sebastian Mueller's Linkedin Analytics