Artificial intelligence is fundamentally reshaping the media landscape, shifting the paradigm from audiences searching for content to content searching for audiences. In this course, you'll explore the sophisticated algorithms driving global platforms like Netflix, YouTube and Spotify, unpacking the mechanics of supervised, unsupervised and reinforcement learning. You'll also examine the critical challenges of algorithmic bias and 'rabbit holes', alongside strategies for reputation management and 'Answer Engine Optimisation' in an age of generative AI.

Machine learning is the engine room of modern media, dictating how content is discovered, consumed and monetised. In this module, you will deconstruct the three fundamental methodologies that drive recommendation algorithms: supervised, unsupervised and reinforcement learning. By examining real-world applications from platforms such as Spotify, Netflix and TikTok, you will learn how 'agents' are trained to maximise engagement, how hidden patterns are discovered in raw data and how organisations balance the need to explore new content against the need to exploit known user preferences.

Moving beyond the theoretical concepts of machine learning, this module examines how major media organisations practically deploy recommendation engines to define the modern user experience. You will explore the specific mechanics of content-based and collaborative filtering, understanding how algorithms determine what content appears in a user’s feed. Through deep-dive case studies of industry leaders like Netflix and YouTube, you will analyse how these systems are engineered to maximise 'stickiness', reduce churn and optimise Customer Lifetime Value using a mix of explicit and implicit data signals.

While recommendation algorithms are powerful engines for engagement, they carry significant risks when optimised solely for attention. In this module, you'll examine the 'rabbit hole' phenomenon, where users are inadvertently led towards extreme content or trapped in filter bubbles. You'll explore the trade-off between algorithmic efficiency and societal well-being and discover the 'algotorial' approach. This hybrid strategy combines the speed of machine learning with human editorial judgement, offering organisations a practical framework to mitigate risk, ensure diversity and maintain safety in content distribution.

In the modern media landscape, an organisation's reputation is increasingly defined not just by general internet sentiment, but by the 'opinions' of large language models (LLMs). This module explores the strategic shift from traditional search engine optimisation to Answer Engine Optimisation (AEO). You will examine how recommendation algorithms perceive brands, the impact of AI inference on public standing, and the practical application of data science in content creation to master the algorithms of platforms like YouTube and Spotify.

This final module consolidates your learning from across the course, summarising the core mechanics of supervised, unsupervised and reinforcement learning, the practical applications of content and collaborative filtering used by platforms like Netflix, Spotify and YouTube, the ethical risks of the 'rabbit hole' effect and strategies for reputation management and creator optimisation in algorithmic systems. Finally, you will apply your knowledge through a peer-reviewed assignment that challenges you to develop a strategic recommendation for how a streaming service could adopt algorithmic recommendation systems, and to analyse the societal impacts of content recommendation algorithms, including the risk of radicalisation.