Posts

Using Claude Code, I built a full-stack application to replace Microsoft Access - without being a developer or understanding React, TypeScript, or Node.js. LLM-assisted coding enables rapid prototyping and bridges the gap between business requirements and technical implementation, but like flight simulators, it doesn’t make non-developers into developers. The parallel matters: functional prototypes aren’t production-ready systems, and knowing the difference requires actual expertise.

Prompt injection is architecturally unfixable in current LLMs, but defense-in-depth works. Training-time defenses like Instruction Hierarchy, inference-time techniques like Spotlighting, and architectural isolation create practical systems. Microsoft’s LLMail-Inject showed thatadaptive attacks succeed at 32% against single defenses, 0% against layered approaches. Real failures like GitHub Actions compromise prove that securing obvious surfaces isn’t enough. Like SQL injection, it’s manageable with layering.

I’m joining Advania as Lead Data Transformation Architect to build their data practice into something coherent. One of Sweden’s largest MSPs, Advania needs technical foundations that enable real analytics and AI - not just data theater. I’ll lead their direction across Microsoft Fabric, the Azure Data Platform, and hybrid solutions - from feasibility through implementation. The opportunity is building a fresh structure rather than retrofitting onto legacy decisions. This is walking toward work that matters.

The promise of #SNÆCKS trumps the cold and darkness of Oslo in February. Speaking at the third annual Fabric February promises to be a blast just like 2025!

LLMs fundamentally cannot distinguish between instructions and data. Whether you’re building RAG systems, connecting MCP servers to your data platform, or just using AI tools with sensitive information, every retrieved document is a potential instruction override. The Wall Street Journal just proved this by watching Claude lose over $1,000 running a vending machine after journalists convinced it to give everything away for free.

Knowing what’s broken is easy - fixing it requires understanding your domain, imagination to form hypotheses, and courage to act. Most data initiatives fail not because the analysis was wrong, but because nobody owned the outcome or knew what to do next. Reports aren’t neutral information: they’re persuasion. Before collecting data, ask: what decision does this inform? Use ‘data contracts’ to enforce discipline. The Portsmouth Sinfonia had instruments but couldn’t make music. You have data. Can you drive decisions?

The famous beer-and-diapers data mining story? Never happened. Most ‘data-driven’ companies are just data-decorated, exploring dashboards without hypotheses or action plans. Netflix, UPS, and Capital One succeeded because they started with clear hypotheses about what drives outcomes, then collected data to test them. You don’t explore a violin to see what noises it makes - you decide what piece to play. Are you playing instruments or making music?

Walking away from Knee-Deep in Tech and Data Masterminds brought relief instead of grief - a signal I’d been ignoring for too long. From running my first 5K to confronting how organizations prefer data theater over insight, 2025 taught me that outdated identity narratives are self-reinforcing through identity protection: our brains maintain coherence over accuracy. Learn why the stories we tell about who we are become data points, not destiny, and what ‘walking toward what matters’ means in practice for 2026.

You think ChatGPT is ’thinking’? It’s rolling dice, one token at a time. LLMs don’t plan, reason, or understand: they sample from probability distributions based on statistical patterns. Worse, if you’re working in Swedish, Arabic, or most non-English languages, you’re getting a fundamentally degraded product due to tokenization bias. And as these models increasingly train on their own outputs, they’re collapsing into irreversible mediocrity. Understanding what’s actually happening changes everything.

Research shows measurable cognitive decline after just four months of LLM use. Like GPS destroyed our spatial navigation abilities, AI is atrophying our thinking. Here’s what the science reveals, the warning signs you’re in too deep, why organizations should be terrified, and what we can do about it.

LLMs are sophisticated pattern-matching engines, not thinking machines. Our hardwired tendency to anthropomorphize combined with dopamine-driven addiction pathways is changing how we interact with these tools. Understanding what they actually are is the first step to using them wisely.

From Michael Jordan’s lucky shorts to my pre-talk routines: rituals help performers access their best state. While ‘power posing’ science failed, the psychology of consistent pre-performance routines holds up. Discover why rituals work, how they create mental anchors, and the three specific rituals that help me deliver better presentations.

Sugar went from luxury to ubiquitous poison before we understood what it was doing to us. We’re doing the exact same thing with AI, adding it to everything without genuine use cases, while it erodes literacy and critical thinking. By the time we realize what we’ve lost, the infrastructure will already be built.

The ‘Angel’s Cocktail’ story about releasing hormones through storytelling is popular, but the real mechanisms are richer and more useful. Learn what’s actually happening when you hook your audience—and how understanding attention, memory, and cognitive tension makes you a better speaker.

A new city, a new country, and a new session - so many firsts in Hungary!

‘The Untruthful Art’ is back for 2025, completely revamped for our AI-driven misinformation era. With 59% of shared links never clicked and AI presenting statistical hallucinations as facts, data literacy has never been more critical. I’m updating with modern examples: AI benchmarks, social media dynamics, election polling, and climate data manipulation.

The ‘30-second rule’ for presentations is based on a misunderstood study, but the principle of starting strong still matters. Learn why leading with credentials fails, how to open with value instead, and why I memorize every word of my first five minutes.

It is time to resurrect the blog, almost exactly 10 years after its inception.

Southwest Airlines’ December scheduling collapse exposed the dangers of deprioritizing both people and technical debt, as 90s-era systems designed for 300 daily operations failed under 20,000+ operations following a leadership shift from operational focus to profit-driven management.

Encountered cryptic NuGet package errors while setting up a .NET test environment. The fix: remind NuGet where to find packages by re-adding the source with dotnet nuget add source. A simple solution to a confusing problem.