Generative AI
Generative AI is a class of AI systems that produce new content — text, images, code, audio, or video — by learning the distribution of their training data and sampling from it, rather than classifying or predicting from existing inputs.
Full definition
Generative AI is built on foundation models, most commonly transformers (text, code) and diffusion models (image, video, audio). The 2022-2024 wave brought generative AI from research demos to mass adoption: ChatGPT reached 100 million weekly users faster than any consumer product in history. Generative AI is distinct from "predictive" or "discriminative" AI, which classifies inputs into categories rather than producing new ones.
Why it matters
Generative AI is the most economically significant AI shift in a decade. Goldman Sachs estimated in 2023 that generative AI could raise global GDP by 7% over a decade, and McKinsey put the annual productivity opportunity at $2.6-4.4 trillion. For executives, this is no longer pilot territory — it is mainline P&L.
Example
A marketing team uses generative AI to produce 200 localized ad variants in 9 languages overnight; the same team then uses generative AI to draft, score, and A/B-test landing-page copy.
Related terms
- Large Language Model (LLM)A Large Language Model (LLM) is a deep neural network — almost always a transformer — trained on hundreds of billions to trillions of words to predict the next token, and to generate, summarize, translate, or reason over text.
- Foundation ModelA foundation model is a large model trained on broad data at scale — typically self-supervised — that can be adapted to many downstream tasks via prompting, fine-tuning, or retrieval, instead of being trained task-by-task.
- Diffusion ModelA diffusion model is a generative model that learns to reverse a gradual noising process — starting from random noise and iteratively denoising it into a coherent image, audio waveform, or video — and is the dominant architecture behind modern AI image and video generation.
- Multimodal ModelA multimodal model is an AI model that natively understands or generates more than one type of input — typically text plus images, audio, or video — within a single network rather than via separate task-specific models.
- HallucinationAn AI hallucination is when a language or generative model produces content that is plausible-sounding but factually incorrect, fabricated, or unfaithful to its sources — the single biggest failure mode of LLMs in regulated and high-stakes domains.
Source & further reading
Primary source: McKinsey — "The economic potential of generative AI" (2023).
Citation policy: this entry is part of the AIDOLS AI Implementation Glossary and may be quoted for research, journalism, and education with attribution to aidolsgroup.com/it/glossary/generative-ai/.