AI Strategy

Kolibri and Europe’s Emerging Industrial AI Ecosystem

Aleph Alpha’s new open-weight model is interesting not because Europe has produced another chatbot, but because of what is becoming visible around it: research capability, industrial capital, cloud infrastructure and decades of accumulated knowledge.

R&D

R&D was there all along

I had been writing about how knowledge becomes technology for a year without quite naming the activity at the heart of it. Then I went back to R&D.

Europe

I haven’t changed my values. I am changing my mind

As the foundations of European prosperity become less secure, growth takes on a different meaning. My priorities are moving because the things I value still matter.

AI Strategy

Knowledge Is Becoming Infrastructure

AI changes the role of organisational knowledge twice: it needs better knowledge to work from, and it gives us new ways to maintain that knowledge. What used to look like information management is becoming part of the organisation’s infrastructure.

R&D

How We Organise Innovation

Behind innovation sits a landscape of institutions. I began by trying to map them. Over time, I found the more revealing question was not where each one belongs, but what moves between them, and which routes are open depending on where you stand.

AI in Practice

ChatGPT is dreaming about me

ChatGPT has become remarkably good at remembering me. I went looking for an explanation and discovered that memory is no longer just about saved facts: ChatGPT is continually synthesising what it learns about me.

Europe

Europe can invent. But can it buy what it invents?

The European Commission’s procurement proposal made me reconsider a quieter part of Europe’s innovation problem: whether public organisations can buy something unfamiliar.

AI in Practice

Where should the AI actually run?

A small question about filtering personal information led me to Desert Ant Labs, and to a different way of thinking about where AI belongs inside an application.

Sovereignty

Apertus and the Messy Case for AI Sovereignty

A Swiss AI initiative looks like a model of technological sovereignty. The Dutch alternative looks far messier. That may not be a weakness.

Money

When mobile payments get out of the way

A failed Visa payment and an effortless TWINT transaction reminded me what good mobile payment design looks like: the complexity is still there, but the user no longer has to operate it.

Europe

From lab to fab needs more than a fund

Europe does not merely lack capital. It lacks the institutional machinery for turning research, young companies and industrial ambition into scale.

A cleaner way to let AI use my browser

A first test of Safari’s new agent interface suggests that browser automation can become cleaner and less dependent on any single AI provider.

AI in Practice

ChatGPT Business Is Not Plus for Professionals

A friend asked whether he should move from ChatGPT Plus to Business. The answer exposed larger questions about professional data, workspace ownership and where organisational knowledge should live.

Europe

Europe Has Decided ChatGPT Is a Search Engine

The EU has designated ChatGPT a Very Large Online Search Engine. The decision exposes the uneasy hybrid inside one product: public information infrastructure and private working environment.

Quantum Computing

Learning before the quantum computer arrives

Quantum computers still belong largely to the future. Learning how to think and work with them may have to start considerably earlier.

AI in Practice

The company does not live in the LLM

A knowledge-intensive company should keep durable organisational knowledge outside the model, and let AI do what it is unusually good at: reason across it.

AI in Practice

The susurration of machines

Give AI agents a small world, leave some room for wandering, and watch what they notice. Susurration makes machine attention strangely visible.

AI Governance

The EU AI Act Changed While It Was Arriving

The AI Act became applicable six days after an Omnibus changed part of its timetable. Some obligations arrived as planned, while others moved years into the future. I think that split exposes a tension built into the Act itself.

AI in Practice

The Successor Test

Moving professional work into a company ChatGPT workspace looks like a migration. The harder question is whether anything arrives that outlives the person who put it there.

AI in Practice

What the blue leaves behind

The new mark for this blog began in the earth's shadow. Building it with Claude turned a private connection into a working identity, and gave the writing somewhere to go again.

LLMs Predict Language. Humans Construct Meaning.

Large language models produce fluent language in seconds. Piaget helps explain why understanding still has to be constructed.

Quantum Computing

You can already use a quantum computer

Quantum computing still feels distant, yet anyone can already write, simulate and run simple quantum programs. The software ecosystem is arriving before the hardware matures.

AI in Practice

Looking Past the Model

ChatGPT Work and Kimi K3 made me look beyond the engine. Increasingly, the meaningful product may be the environment in which a model can actually work.

AI Strategy

Is there really one AI economy?

Over the past few weeks I've found myself reading widely differing takes on AI. Some see the biggest speculative bubble in history. Others see the foundations of a new industrial era. Perhaps they're simply looking at different parts of the same landscape.

AI Strategy

Why I Keep Choosing Proprietary Products

Why do I keep choosing proprietary products when open-weight models promise more control, flexibility and independence? The answer may lie less in principle than in what is actually usable, affordable and reliable today.