45 Entries · 17 Letters
Glossary
A reference of terms used in essays on this site. Definitions are short; each entry links to a longer note and to the writing in which the term appears.
A
- Agentic AIAI systems capable of autonomous action—perceiving their environment, making decisions, and executing multi-step tasks with minimal human intervention.
- AI RegulationLegal frameworks, policies, and standards governing the development and deployment of artificial intelligence systems.
- AI SafetyResearch and practices aimed at ensuring AI systems behave as intended and don't cause unintended harm, from near-term risks to existential concerns.
- AI TransparencyThe practice of making AI systems understandable—disclosing how they work, what data they use, and how decisions are made.
- AI WatermarkingTechniques to embed invisible markers in AI-generated content to enable identification of its synthetic origin.
- Algorithmic BiasSystematic errors in AI systems that create unfair outcomes, typically reflecting biases present in training data or design choices.
- APIApplication Programming Interface—a set of rules allowing software systems to communicate, enabling developers to access AI capabilities programmatically.
C
- ChatbotSoftware that simulates conversation with users, ranging from rule-based scripts to sophisticated AI assistants powered by LLMs.
- ComputeThe computational resources—processing power, memory, and infrastructure—required to train and run AI models.
- Content AuthenticityTechnologies and standards for verifying the origin, history, and integrity of digital content in an era of AI-generated media.
- Content ModerationThe practice of monitoring and filtering user-generated content to enforce platform policies and legal requirements.
- Context WindowThe maximum amount of text an LLM can process at once—the model's working memory that limits how much it can 'see' in a conversation.
- CopyrightLegal protection granting creators exclusive rights to their original works, now central to debates over AI training data and outputs.
D
- Dark PatternsDeceptive user interface designs that manipulate users into actions they didn't intend, from unwanted purchases to privacy violations.
- Data PrivacyThe right of individuals to control how their personal information is collected, used, and shared by organizations and AI systems.
- DeepfakeSynthetic media created using AI to realistically depict someone saying or doing something they never actually did.
- Diffusion ModelA generative AI architecture that creates images by learning to reverse a gradual noising process, powering systems like Stable Diffusion and DALL-E.
- DisinformationFalse or misleading information deliberately created and spread to deceive, distinct from misinformation which may be shared unknowingly.
E
- EmbeddingA numerical representation of text, images, or other data as vectors, enabling AI to measure similarity and meaning mathematically.
- End-to-End EncryptionA security method where only the communicating parties can read messages, with no access possible for service providers or intermediaries.
- EnshittificationThe gradual degradation of online platforms as they shift value from users to advertisers and shareholders, ultimately harming everyone.
F
- Facial RecognitionAI technology that identifies or verifies individuals by analyzing facial features, raising significant privacy and civil liberties concerns.
- Fine-TuningThe process of further training a pre-trained AI model on domain-specific data to specialize its capabilities for particular tasks.
G
H
I
L
M
- Machine LearningA subset of AI where systems learn patterns from data rather than being explicitly programmed with rules.
- MisinformationFalse or inaccurate information spread without intent to deceive, distinct from disinformation which is deliberately misleading.
- Model AlignmentThe challenge of ensuring AI systems behave according to human values and intentions, avoiding harmful or unintended behaviors.
- Multimodal AIAI systems that can process and generate multiple types of content—text, images, audio, and video—within a single model.
N
O
P
R
- RAGRetrieval-Augmented Generation—a technique that enhances LLM responses by first retrieving relevant information from external sources.
- Recommendation AlgorithmAI systems that predict what content, products, or connections users will find relevant, powering feeds across social media, streaming, and e-commerce.
- RLHFReinforcement Learning from Human Feedback—a technique to align AI behavior with human preferences by training on human evaluations.
S
- Social TVThe integration of social media with television viewing, enabling real-time audience engagement and second-screen experiences.
- Surveillance CapitalismAn economic system where personal data is extracted and commodified for behavioral prediction and manipulation, often without meaningful consent.
- Synthetic MediaAny media—images, video, audio, or text—generated or substantially modified by AI rather than captured or created by humans.
T
- Text-to-ImageAI systems that generate images from written descriptions, transforming text prompts into visual content.
- Text-to-VideoAI systems that generate video content from text descriptions, representing a frontier in generative AI.
- Training DataThe dataset used to teach an AI model patterns and behaviors—the examples from which it learns.
- TransformerA neural network architecture using self-attention mechanisms, forming the foundation of modern LLMs like GPT and Claude.