In every CAS class and client session, one question keeps popping up:
"What should I really know to work well with AI?"
Hint: it’s more than just prompting ChatGPT.

Successful AI comes from a solid, connected skillset…

🙂In 2023, Noy & Zhang (Science, Jul 2023) reported that a GPT-based copilot made knowledge workers «faster, better, happier».

Since then, chatbots have come a long way. They need to be fast. Accurate. Need to handle knowledge base updates…

Imagine inviting friends for dinner. You head to a shop where everything is beautifully displayed, labeled with farmer names, organic tags, and allergy warnings. Behind the scenes, someone ensures freshness, removes spoiled produce, and keeps the shelves stocked…

Every now and then, I'm asked to review a data platform or data organization. These reviews can feel frustrating. Or become real opportunities for growth.

✅ Objective criteria matter:
- Technology choices
- Architecture design…

What do you get when you combine medical imaging, AI, and a 2000-year-old mystery? A masterclass in cross-domain innovation.

We all know Pompeii, the city frozen in time by volcanic ash…

Remember your childhood treasure box, where you kept your treasures (candy bars 🍫, pictures of your heroes 📸) safe from prying eyes. Only those with a key could peek inside. Many companies still handle data this way: defining boxes (like salary data) and tightly controlling who sees inside.

But data isn’t static or neatly siloed…

The turkey thinks the farmer loves him. Every day, it gets fed. Every day, it predicts: “Tomorrow will be another good day.” Until Thanksgiving (at least in some places). That’s the risk of taking things for granted. Recently, the stock market has been moving heavily…

Remember creating cheat sheets back in school? Tiny notes packed with formulas, concepts, and shortcuts. The real value wasn’t having them during the test; it was the process of making them: condensing complex ideas into essentials.

That’s the magic of learning: Compression. Understanding what truly matters…

Imagine inviting friends for dinner. You head to a shop where everything is beautifully displayed, labeled with farmer names, organic tags, and allergy warnings. Behind the scenes, someone ensures freshness, removes spoiled produce, and keeps the shelves stocked…

Trend: Incremental, rapid iteration combined with AI-driven discovery is becoming a dominant innovation strategy globally—especially visible in the biotech sector.

"In recent months, China’s progress in artificial intelligence has stunned the world. A quieter…

Back when I taught physics to medical students, I learned that the right analogy could make even the toughest concepts click. AI is no different.  Let’s break it down with four fun comparisons:

🎿 Gradient Descent → Finding the Best Fall Line
A skier instinctively follows the steepest downhill path. AI does the same—adjusting parameters step by step to minimize error…

A photo of a crowded event—how many people are in it? A phone call transcript—who spoke when? Not long ago, these were “unstructured” data points. Today, AI extracts faces from images and timestamps from conversations … yet many teams still draw their data transformations in visual tools—then jump to a “free-hand SQL” section when complexity strikes.

“The future is already here—it’s just not evenly distributed.”
– William Gibson

Everett Rogers’ Diffusion of Innovations Theory (1962) reminds us that breakthroughs spread in stages—early adopters get the benefits first while many lag behind. Today, three vivid examples highlight this uneven progress:

A fly zips through the air with just 150,000 neurons, consuming only microwatts of energy. It stabilizes mid-flight, dodges obstacles, and reacts faster than you can swat. Meanwhile, large AI models require vast computational resources, scaling energy consumption with every generated token.

Are we building intelligence the wrong way?

Did you learn to swim without water? 🏊‍♂️ Building data solutions without real data is just as impossible!

For years, IT has relied on the Dev → Test → Preprod → Prod model. While this works for classic software, it falls short for data and AI. Test data rarely captures the real-world complexity created daily in fast-paced, unpredictable business processes. Only live data reveals the true patterns needed to build robust pipelines and train accurate models. 📊

Recently, I strapped on the same glucose monitor people with diabetes rely on, all to learn more about my food patterns. What surprised me: how stable my blood sugar remained—even overnight. I assumed it would steadily drop in the wee hours, but our body is great at tapping into reserves.

It turns out blood sugar and body temperature aren’t the only strictly regulated markers. A new study in Nature…

For centuries, many believed the Earth was fixed, continents never shifted, and diseases came from “bad air.” Then science stepped in—with simple observations like falling apples, fossil records, and microscopic bacteria—to set the record straight.

A modern example is NASA’s Parker Solar Probe. It’s flying closer to the Sun than ever (unlike Icarus, it’s doing just fine) to uncover hidden truths.

Companies recognize the pivotal role of data and AI in driving success. This realization challenges data systems to process more sources, meet rising expectations, and deliver faster decisions.

The key? Balancing stability for consistent, reliable outcomes with agility to adapt and innovate.

The solution lies in a platform built on four focus areas…

TechCrunch’s list of disruptive startups for 2024 puts a spotlight on innovation. Here are topics which particularly resonate with me - proof of what’s possible when data, AI, and brilliant engineering come together:

🏥 Healthcare
Abridge transforms doctor-patient conversations…

Es beginnt alles mit einer Idee.