- Reading time: 3 minutes
- Price: Free download
- Published: 3rd October 2026
- Word count: 615 words
- File format: Text
Personal statement example
Most of my shifts as a self-checkout supervisor involve overriding a machine. The scales decide a bag of loose carrots weighs too much, or the camera judges a customer too young to buy paracetamol, and a light flashes until I walk over. After eighteen months I have noticed patterns. Certain customers, often older people or those buying a full week's shop with children in tow, trigger interventions far more than others, and they tend to apologise as though the error were theirs. The system is not malicious, but it quietly distributes small inconveniences and small embarrassments unevenly. That observation, ordinary as it is, is a large part of why I want to study the ethics of AI at postgraduate level.
My undergraduate degree was in Philosophy, and I came to these questions through moral and political philosophy rather than computing. A second-year module on theories of justice gave me a vocabulary for distribution and desert, but I wanted to test it on something concrete. My dissertation examined automated CV screening and asked what it would mean for such a tool to be fair. I worked through the literature showing that several common statistical definitions of fairness, such as equal error rates across groups and calibration of scores, cannot generally all be satisfied at once when groups differ in their underlying rates. My argument was that this is not merely a technical trade-off for engineers to settle, because choosing which definition to prioritise is itself a moral and political decision that should be made openly. My supervisor pushed me to engage with objections from a broadly consequentialist position, and rewriting that chapter taught me more than any other part of the project.
Reading Virginia Eubanks's Automating Inequality alongside the dissertation widened my focus. Her accounts of automated eligibility and risk-scoring systems in American public services showed me that the harms of these tools often arise from how they are embedded in institutions, not only from flaws in the models. I want to develop this institutional perspective further, particularly around public-facing services in the UK.
To understand the technical side well enough to argue about it responsibly, I completed a free online introduction to Python and worked through a beginner's tutorial on training a simple classifier on a public dataset. I am under no illusion that this makes me a data scientist, but it means I can follow what a confusion matrix shows and ask better questions about where training data comes from.
Outside work, I volunteer fortnightly at a library drop-in session helping people with online forms, mostly benefit applications, council tax queries and GP booking systems. Much of the job is patient explanation, and I have become good at it: I can usually help someone renew a parking permit or upload a document in a single session, and several regulars now come to me by name. The sessions have also shown me how much depends on interface choices that designers probably never imagined being difficult. I also sing tenor in a community choir, which has nothing to do with algorithms but has made me comfortable rehearsing slowly and taking direction.
I am applying for an MSc because I want structured training in applied ethics, governance and the social study of technology, and the chance to research these questions more rigorously than my dissertation allowed. In the longer term I would like to work in policy or public-sector assurance, helping organisations decide when automation is appropriate and how its effects should be monitored. I bring philosophical training, a practical grasp of how people actually meet automated systems, and the habit of noticing who is being asked to apologise for the machine.