Thinking out loud.

Building an agent is now cheap enough that the person staring at the problem can prototype a fix themselves. But 'should they build it' hides three different questions, and only one of them belongs to the builder.

Every model release quietly resets part of what you know. The prompt that pulled a brilliant answer out of 4.2 can wreck the response in 5.5. If execution is no longer the bottleneck, what is?

The current excitement around AI loops and workflows isn't new — we have always worked in loops. What's changed is the speed. What hasn't changed is what makes them actually learn.

You step into finance, HR, customer support, marketing, operations — and the silence is deafening. This is the gap nobody talks about. AI adoption can't be an engineering-only conversation.

The line between a 'technology company' and 'everyone else' is gone. It didn't blur — it collapsed. Your competitors aren't waiting for a digital transformation roadmap. They're shipping.

Solutions are easy. Problems are hard. When the barrier to building goes down, the risk of building the wrong thing goes up. The real competitive advantage isn't the ability to build quickly — it's the ability to choose wisely.

AI does not replace the work that matters. It replaces the work around the work that matters. If you bring expertise and judgement, these tools will multiply you. If you bring vagueness and hope, they will multiply that too.

If it feels like the AI landscape is changing weekly — it's because it is. The pace of change means we need a different mindset. We need to stay close to the tools, test continuously, and build curiosity into the way we work.

Everything is easy when you already know how to do it. But starting from scratch? That's something very different. AI doesn't become powerful when only a few master it — it becomes powerful when many feel safe enough to try.

Digital transformation was often for the few. AI transformation requires everyone. Ten years later, the same trainee group met again — and this time, the conversation is entirely different.

Organizations have been defined as groups of people working together. But as AI agents enter everyday workflows, that definition is already outdated. We're moving to hybrid teams — and most organizations aren't ready.

To build something great, you must first be willing to be truly, unapologetically bad at it. I'm building 30 agents and stepping straight into unfamiliar territory.

The biggest bottleneck in an AI-driven environment is rarely the technology — it's us. Our habits, our literacy, our comfort zones. The journey starts with you.

Hashtags do not save the world — knowledge and your actions do. Applause creates feelings. Action creates change.

The walls separating doers from dreamers are crumbling. With AI tools, your ability to innovate isn't defined by your technical pedigree — it's powered by your curiosity.

Only 19% of Nordic white-collar workers use GenAI weekly, compared to 61% globally. The Nordics risk falling behind — but it doesn't have to be that way.

My first five days at twoday, I made AI my onboarding buddy — from getting to know my team to building agents and running product assessments. Here's what I learned.

Dogfooding means using your own product every day. As consultants, it's how we build genuine empathy and credibility — by experiencing the same frustrations our clients do.

Around 4.5 years ago, my partner and I started dreaming about a house on a lot near my parents. What followed was less a construction project and more a product journey.

We have centuries-old rites of passage for life's milestones — but almost none for the products we build and kill. Maybe we should.

Sharing your doctor appointments and kindergarten pickups in your work calendar might sound like oversharing — but it could be one of the most powerful things you do for your team culture.