Data fit for AI
Data foundations, data ethics and algorithmic fairness.
- Level
- Bitesize
- Learning time
- 1 hour 15 minutes
- Where
- Strategy and the future
- Price
- £39
When you step up to a higher level, what you paid in the last 12 months for lower-level courses inside the bigger course comes off its price, in proportion to how much of each is inside it, whichever topic it's from. Courses at the same or a lower level don't give credit against each other. A Focus course or above always costs at least £15. Read the full terms
Certificate of Completion
Overview and Outcomes.
Overview
Every AI output inherits the strengths and flaws of the data and culture beneath it, and when data fails, people are harmed. The pairing opens with data foundations: five tests of data quality, the idea that a model learns only from the history its data contains, who is missing from your data, and the 2020 A level grading algorithm and Robodebt as cases of real harm. It then turns to data culture and ethics, drawing on Donella Meadows on information flows and the Government Digital Service's Data and AI Ethics Framework, and shows how bias scales through the DWP's fairness analysis of its Universal Credit advances model.
What you'll develop
In this Bitesize, you will develop the ability to:
- Explain why AI outputs inherit the quality of their data and why named ownership of data sets matters.
- Describe how organisational culture shapes what data is allowed to say, and identify the features of a healthy data culture.
- Apply five tests of data quality to a proposed AI use, checking it is accurate, complete, timely, consistent and representative.
- Apply the principles of the Data and AI Ethics Framework, treating fairness as a continuing choice rather than a box ticked once.
- Identify who is missing from your data and plan the steps, from inventory to monitoring, that get data ready for AI.
- Distinguish a data-driven from a data-informed stance, and judge which keeps people accountable for decisions in your setting.
The Modules.
1 hour 15 minutes of learning.
- 06Data Foundations: No AI Without Good DataEvery AI output inherits the strengths and flaws of the data beneath it. This module sets out five foundations beneath any AI system, five tests of data quality, and the idea that a model can only learn from the history its data contains. Through the 2020 A level grading algorithm and Robodebt it shows how data problems harm people, then covers data protection essentials, who is missing from your data, and practical steps to get data ready.39 min
- 16Data Culture, Ethics and Algorithmic FairnessData is made by people, so its quality depends on how people feel about producing, sharing and hearing it. This module covers an Avanti view of data maturity, five features of a healthy data culture, Donella Meadows on information flows, the Government Digital Service's Data and AI Ethics Framework, and how bias scales inside automated systems, including the DWP's own fairness analysis of its Universal Credit advances model.39 min