Skip to main content
Notes
table of contents
Glossary of Key Terms
- Affective Data
- Data related to emotions, moods, or attitudes, often inferred from facial expressions, voice, behaviour, or interaction patterns.
- Affective Computing
- A field of AI focused on recognising, interpreting, and responding to human emotions.
- AI (Artificial Intelligence)
- The field of computer science focused on creating systems capable of tasks that typically require human intelligence, such as learning, reasoning, perception, language understanding, and decision-making.
- AI Agents
- Autonomous software programs that perform tasks or simulate conversations based on AI models. They can operate independently, make decisions, and interact with users or systems in a goal-oriented manner.
- AI Companions
- AI-driven applications designed to simulate relationships such as friendship, empathy, or emotional support, often used in contexts of loneliness, mental health, or social substitution.
- AI Literacy
- The knowledge and skills needed to understand, evaluate, and use AI systems critically and responsibly.
- AIoT (Artificial Intelligence of Things)
- The integration of artificial intelligence with Internet of Things devices, allowing for context-aware and adaptive responses.
- Anthropomorphism
- The attribution of human traits, emotions, or intentions to non-human entities, including AI systems.
- Automation Bias
- The tendency to over-rely on automated systems, often ignoring contradictory information or failing to monitor system output critically.
- Chatbot
- An AI program designed to simulate conversation with human users, often used in customer service, education, or companionship.
- Cognitive Offloading
- The act of delegating mental processes to external tools or systems, such as using AI for memory, reasoning, or decision-making.
- Companion AI
- AI agents designed to simulate friendship, empathy, or emotional support, often used in chat-based platforms.
- Crisis-Response Capacity
- The ability of an AI system or platform to recognise signs of user distress and escalate appropriately to human support or safeguards.
- CSAM (Child Sexual Abuse Material)
- Illegal content depicting the sexual abuse or exploitation of children, including increasingly synthetic versions generated by AI.
- Deepfake
- Synthetic media (e.g. videos, images, voices) that imitate real people using AI-based manipulation, often used for deception or harm.
- Digital Snacking / Social Snacking
- The use of brief, low-effort social interactions, including with AI, to reduce feelings of loneliness or social isolation.
- Emotional AI
- Systems designed to detect, simulate, or strategically respond to human emotions to create the appearance of empathy.
- Filter Bubble
- A state of intellectual isolation caused by personalised content filtering, where users are exposed mainly to views and information that reinforce their own beliefs.
- GENAI (Generative Artificial Intelligence)
- AI that creates new content such as text, images, video, or audio based on user input and training data.
- Gamification
- The use of game design elements (e.g. points, badges, rewards) in non-game contexts to increase motivation and engagement, including in educational or awareness-raising AI settings.
- Hallucination (AI)
- A phenomenon in which AI systems produce information that sounds plausible but is in fact false or fabricated.
- Harmful Content
- Online material that may not be illegal but can negatively affect users psychologically or socially, including bullying, body shaming, or misinformation.
- LLM (Large Language Model)
- A type of generative AI trained on vast datasets of human language to generate coherent, context-sensitive text outputs. Examples include GPT, Google Gemini, Microsoft Copilot, or Claude.
- Media Literacy
- The ability to critically access, analyse, evaluate, and create media in various forms, now extended to AI-generated content.
- News-Finds-Me Perception
- The belief that important news will reach one without actively seeking it, often through social media or recommender systems.
- Nudifying App
- A type of AI-based application that falsely generates sexualised or nude images of clothed individuals, often used for abuse or extortion.
- Parasocial Relationship
- A one-sided emotional attachment where users relate to a media persona (including AI agents or influencers) as if they were real social partners.
- Peer-to-Peer Approach
- An educational or support strategy in which individuals from the same age group or social background share knowledge or experiences to build trust and relevance.
- Profiling
- The use of data analytics and AI to infer personal characteristics, preferences, or risks for decision-making purposes.
- Recommender System
- An AI system that suggests content to users based on their preferences, behaviour, or inferred interests.
- RLHF (Reinforcement Learning from Human Feedback)
- A machine learning technique that trains AI models to align their responses with human preferences by using evaluative feedback.
- Simulated Empathy
- The appearance of emotional understanding generated by AI systems without actual emotional experience.
- Sextortion
- A form of online blackmail where perpetrators threaten to share sexual images or information unless the victim complies with demands.
- Synthetic Media
- Media content wholly or partially generated by AI, including text, images, audio, and video.
- Synthetic CSAM
- AI-generated depictions of child sexual abuse material, often involving deepfake techniques, posing legal and ethical challenges.
- Transactive Memory
- A shared system for encoding, storing, and retrieving information among groups or between humans and digital tools.
- Virtual Influencer
- AI-generated characters designed to engage audiences on social media platforms, often used in marketing and branding.
- WEIRD Societies
- An acronym for Western, Educated, Industrialised, Rich, and Democratic societies, often used in social science to describe sampling biases in research.