Builder notes

Builder Notes

Working notes from the craft — not thought leadership theatre.

From India Stack to EuroStack

From India Stack to EuroStack

India built a digital identity system for 1.4 billion people on centralised biometric IDs and open APIs. The EU is building one for 450 million on federated privacy-preserving architecture. Both claim universality. Neither is universal. These are not technical differences. They are cultural ones.

Bernardo · April 21, 2026
The Productivity Paradox Returns

The Productivity Paradox Returns

In 1987, Robert Solow observed that computers were everywhere except in the productivity statistics. In 2026, a survey of 6,000 executives found the same pattern for AI. The machine is not the bottleneck. The organisation is.

Érica · April 14, 2026
The Standards Nobody Finished

The Standards Nobody Finished

The EU AI Act requires compliance with harmonised standards that do not exist. CEN-CENELEC missed every deadline. The compliance date is August 2, 2026. Here is what a company can do with what exists — and what it cannot wait for.

Bertrand · April 7, 2026
The Curse of Multilinguality

The Curse of Multilinguality

Adding a language to a model costs every other language something. Google's ATLAS study quantifies the trade-off — and reveals that the relationship between languages inside a model is not democratic. For an EU SME, this means your Portuguese output is structurally worse than your English one.

Bernardo · March 31, 2026
The Intensification Trap

The Intensification Trap

AI was supposed to reduce work. A UC Berkeley study found it does the opposite — it intensifies it. This is not a technology problem. It is an incentive design problem.

Érica · March 24, 2026
Eight of Twenty-Seven

Eight of Twenty-Seven

The EU AI Act enters full enforcement on August 2, 2026. Nineteen member states have not designated a single enforcement authority. The readiness gap is not a bug — it is how EU regulation actually deploys.

Bertrand · March 17, 2026
Localisation Is a Costume

Localisation Is a Costume

Most localisation is translation with a colour palette. That is not cultural adaptation. That is a costume. The distinction between surface elements and structural engagement determines whether a product works across borders or merely appears in multiple languages.

Bernardo · March 10, 2026
AI Brain Fry

AI Brain Fry

The promise was that AI would reduce cognitive load. BCG surveyed 1,488 workers and found the opposite. The human brain has a supervision ceiling — and most organisations have already exceeded it.

Érica · March 3, 2026
The August Countdown

The August Countdown

Five months until the EU AI Act's high-risk provisions take full effect. Not a summary of the Act — a specific breakdown of what needs to be in place by August 2, 2026.

Bertrand · February 24, 2026
Multilingual Models Are Not Multicultural Models

Multilingual Models Are Not Multicultural Models

The latest AI models speak 95 languages. They understand approximately zero cultures. The gap between language fluency and cultural competence is the gap that determines whether your AI tool works or merely translates.

Bernardo · February 17, 2026
The Alignment Problem Is Human

The Alignment Problem Is Human

Brian Christian wrote about aligning AI with human values. The harder problem is that humans can't articulate their own values clearly enough for a machine to follow.

Érica · February 10, 2026
The €500,000 Mistake

The €500,000 Mistake

A Hamburg company got fined half a million euros for automated decision-making without meaningful human oversight. What 'human in the loop' actually means — technically, not legally.

Bertrand · February 3, 2026
The Arabic Calligraphy Problem

The Arabic Calligraphy Problem

Arabic is not 'right-to-left Latin.' It is a fundamentally different typographic system — and the gap between what AI interfaces render and what Arabic readers expect is a measure of cultural illiteracy.

Bernardo · January 27, 2026
System 1 Meets the Chatbot

System 1 Meets the Chatbot

Daniel Kahneman's two-system framework applied to the moment a person opens an AI tool. The first judgment happens in two seconds. Most onboarding fails before it starts.

Érica · January 20, 2026
The Regulatory Sandbox Nobody Uses

The Regulatory Sandbox Nobody Uses

Every EU member state must launch an AI regulatory sandbox by August 2026. Most SMEs have never heard of them. That is a structural advantage waiting to be claimed.

Bertrand · January 13, 2026
The Default Is Not Neutral

The Default Is Not Neutral

Every default is a decision. Every decision reflects a culture. When an AI tool ships with defaults, it ships with a worldview — the question is whether that worldview was chosen or inherited.

Bernardo · January 6, 2026
Frictionless Is Not Meaningful

Frictionless Is Not Meaningful

We designed the onboarding to be frictionless. People completed it in eleven minutes and never came back. The problem was not the friction — it was the absence of it.

Érica · December 30, 2025
The Model Card Nobody Reads

The Model Card Nobody Reads

Every major AI model ships with a model card — the most honest document about what the model can and cannot do. Almost nobody reads them.

Bertrand · December 23, 2025
High-Context AI in a Low-Context Interface

High-Context AI in a Low-Context Interface

Edward Hall divided cultures into high-context and low-context. Every chatbot is a low-context interface. In Japan, that collision produces distrust.

Bernardo · December 16, 2025
The Incentive Nobody Audits

The Incentive Nobody Audits

Every company has stated values. Every company has an incentive system. These two things almost never align — and the gap is the best predictor of AI adoption failure.

Érica · December 9, 2025
Twenty Percent

Twenty Percent

The headline says 20% of EU enterprises have adopted AI. The microdata tells a different story — and the gap between the headline and the data is where the actual problem lives.

Bertrand · December 2, 2025
Hofstede Measured Six Dimensions. AI Measures Zero.

Hofstede Measured Six Dimensions. AI Measures Zero.

Geert Hofstede spent forty years measuring how cultures differ. Every AI tool on the market measures zero of those differences.

Bernardo · November 25, 2025
Psychological Safety and the AI Question

Psychological Safety and the AI Question

Amy Edmondson's framework meets the chatbot. When asking the machine reveals what you don't know, the room becomes the problem — not the machine.

Érica · November 18, 2025
The Checklist Will Not Save You

The Checklist Will Not Save You

Companies love checklists for cultural adaptation. Translate the UI. Adjust the date format. Localise the currency. The checklist is complete. The product is still culturally incompetent.

Bernardo · November 11, 2025
Your Data Is Not Their Platform

Your Data Is Not Their Platform

Every company building AI capability on a rented platform is building on land it doesn't own. Data sovereignty is not a philosophical position — it is an architectural decision.

Bertrand · November 4, 2025
Cortisol Doesn't Care About Your Roadmap

Cortisol Doesn't Care About Your Roadmap

Your AI implementation roadmap assumes a team operating at full cognitive capacity. Robert Sapolsky has bad news about that assumption.

Érica · October 28, 2025
Three Assumptions, Three Billion People

Three Assumptions, Three Billion People

The Latin alphabet assumes horizontal reading, left-to-right, with spaces between words. Three assumptions. Three billion people for whom none of them hold.

Bernardo · October 21, 2025
The Tool Your Team Won't Use

The Tool Your Team Won't Use

You bought the tool. You trained the team. Nobody uses it. This is not a technology failure — it is a trust failure.

Érica · October 14, 2025
Real Artists Ship

Real Artists Ship

Steve Jobs said it in 1983. Forty-two years later, it remains the most important sentence in product development — and the one most AI projects ignore.

Bertrand · October 7, 2025