Introduction
Kwame Dorn first became interested in algorithms because of silence. Young people he worked with sent applications through online portals and received nothing back. No rejection, no explanation, just a page that changed status.
Kwame lives in Hamburg and advises organizations that want to use AI in personnel, education or social administration. His starting point is youth work, not a technical lab.
Story of the Path into AI
Kwame wanted to know whether automated preselection was reinforcing the exclusions he saw every week. He did not come from computer science and had to learn how to make technical criticism specific enough that companies would not dismiss it as mood or politics.
He studied fairness metrics, anti-discrimination law and participatory workshop methods. His first project was a survey of young people about digital application portals and the exclusions they experienced. The results were not neat. Some harms came from algorithms, others from unclear forms, missing feedback or school histories that systems treated as simple signals.
That complexity became his argument. If AI is used in social contexts, affected people must help define the test cases.
Current Work
Today Kwame works with programme providers, employers and public projects. In one funding programme he insisted that young applicants help write test scenarios. The team discovered criteria that disadvantaged non-linear school biographies. The screening process was changed, and borderline cases received human review.
Kwame does not claim that fairness can be solved in a dashboard. Metrics matter, but they come after listening. A system that looks fair in aggregate can still make a particular life harder to explain.
Personal Advice
“Fairness does not emerge in the dashboard. It begins when affected people sit at the table,” Kwame says. His advice to technical teams is to invite contradiction early, before the model has already shaped the process.
Key Facts
Age and place: 29, Hamburg.
Background: social work, community perspective, non-technical entry.
Entry into AI: survey on digital application portals and experienced exclusions.
Focus today: algorithmic fairness.
Typical tools: fairness audits, workshops, qualitative interviews.
Werkstattnotiz
Kwame keeps anonymized application stories in a folder labelled “not an outlier yet.” He dislikes how quickly rare cases are dismissed. His current work asks whether a system can show statistical performance without making unusual lives pay the price for neat categories.