Anti-Hype AI Dictionary
Confused by over-hyped, overlapping, and obtuse AI buzzwords? That’s why I’m working on the Anti-Hype AI Dictionary, offering simple, non-technical definitions of the AI phrases being tossed around the web.
Artificial Intelligence (AI): An umbrella term, encompassing a range of systems designed to “perform tasks that typically require human intelligence”. During this current wave of generative AI, the term ‘AI’ colloquially most often used in reference to generative AI systems (see below) specifically.
Machine Learning (ML): A subset of AI, focused on creating algorithms that learn patterns from data rather than following explicitly programmed rules. Although the term ‘Machine Learning’ has fallen out of favor in recent years, current Generative AI and Discriminative AI techniques are subsets of ML.
Generative AI (GenAI): AI systems designed to create new content (text, images, video, code, etc.) based on patterns learned from training data.
Discriminative (Predictive) AI: Systems that analyze existing data to make predictions about future events or unknown values. Most commonly discriminative AI is used to classify data into one of a set of pre-defined categories or predict a numerical value.
Deep Learning and Neural Networks: An approach to machine learning using multi-layered neural networks loosely inspired by the human brain's structure. Neural networks in some form have been around since the 1940’s, and the term “Deep Learning” is used to describe larger and more complex neural networks.
Transformer Models: Neural network architectures that process sequential data using ‘attention’ mechanisms to identify relationships between elements regardless of their position in the sequence.
Foundation Model: Large AI models trained on vast datasets that can be adapted to many different tasks.
Frontier Models: Foundation models that “exceed the capabilities currently present in the most advanced existing AI models”. Essentially these are the newest, fanciest foundation models. I don’t think this term has much meaning beyond hype.
Large Language Models (LLMs): A subset of foundation models specifically trained on text data to predict and generate human language.
AI Agents: Software systems that can perceive their environment, make decisions, and take actions to achieve specific pre-defined goals. The term is often used loosely to describe any automated system with minimal autonomy, creating confusion about capabilities.
Agentic AI: AI systems designed to act independently to accomplish tasks with minimal human supervision. Agentic AI is frequently promoted as truly autonomous, although it is a very new application of the technology, and the jury is still out on how well these systems really work without supervision. In my view, Agentic AI is on the far end of a continuum with basic automation software, with AI Agents somewhere in the middle.
Artificial Narrow Intelligence (ANI): AI systems designed for specific tasks within a limited domain. In the context of interdisciplinary understandings of intelligence, this might be considered “competence” rather than “intelligence”. The term is sometimes used dismissively, ignoring that even "narrow" systems can be highly sophisticated and valuable.
Artificial General Intelligence (AGI): A hypothetical AI system with human-level cognitive abilities across diverse domains. This term is essentially just a rebrand of the original term “AI” as it was understood by John McCarthy in the 1950’s. When “AI” techniques failed to meet expectations and only produced narrow intelligence, the term ‘AGI’ emerged to became the new north star.
AGI is often often presented as being on a continuum with ANI, although this assumption lacks theoretical or empirical support.
Artificial Super Intelligence (ASI): A hypothetical AI system with capabilities far exceeding human intelligence.