I keep a list of things I haven’t yet explained.It grows in the years when the work goes well.The questions rarely stay in the field where they began.
- 003
Does making reasoning visible improve the decision, or only the record of it?
↗ Selected Work - 018
What is the feeling of being right actually tracking?
↗ Research - 047
Why does a reorganisation everybody agrees with sometimes produce months of quiet refusal?
- 112
If roadmaps rarely predict the future, what are they actually for?
- 131
When does (AI) making a decision easier make the judgment worse?
↗ Research
Gustav Béen
Fifteen years in product and transformation, working across digital products, organisational change, and the decisions that connect them.
Alongside that, I maintain a small independent practice exploring many of the same questions at a human scale. The work now moves between product, writing and research, depending on where those questions lead. Currently writing a dissertation on how AI affects confidence and judgment in professional decision-making under uncertainty. Currently based in Málaga, Spain.
Current Threads
Most conversations don’t fail because people disagree. They fail when both people become more certain about the other’s intention than curious about what is happening between them.
J. G. March, “Exploration and Exploitation in Organizational Learning” (1991). Read for a thesis about individuals; explains most of the organisations I’ve sat inside as well.
We built one SDK so every team could ship the same look, the same components, the same way. I didn’t understand it then, but what we were actually asking them to give up wasn’t code – it was the one thing every team gets to decide for itself, once everything else has been centralised.
When people look back on 2026, I don’t think they’ll say the limit was artificial intelligence. I think they’ll say it was organisational intelligence.
Does the expressed confidence of an AI-generated recommendation influence the calibration (rather than just the accuracy) of professionals’ own judgments under uncertainty?
A decision-journal tool using an LLM to help decision-makers log decisions in the moment and later analyse patterns in their own reasoning – where their confidence diverged from outcomes, where cognitive biases likely crept in.
World-building for a science fiction novel, working title Agnosoma. Untouched for ten months. Filed here so it stays honest.
The hardest part is rarely finding words for what someone feels. It’s discovering what becomes possible once they finally have them.