Top news
Investing in the public benefit OfficialAnthropic NewsOct 18Popular Musk Pledges SpaceX Shares to ‘Every Kid in America’ WebBloomberg TechnologyOct 19Popular Websites are closing their doors to AI agents WebThe Verge AIOct 17 This is the new ChatGPT Voice, powered by GPT Live OfficialOpenAI on YouTubeOct 18 How AI giants will reshape global philanthropy WebAxios AIOct 19 Gemini 4 Argon Really Surprised Me WebTwo Minute PapersOct 19 Google Gemini 4 agents just beat Minecraft with no human intervention WebEngadget AIOct 19 Who could it be? 🧐 OfficialGoogle on YouTubeOct 18 Anthropic gives outside AI watchdog “employee-like” access WebArs Technica AIOct 18 What are self-replicating ‘AI worms’, and why is Anthropic worried about them? WebThe Guardian AIOct 18
GlossaryModelsNew to AI

Generative models

Also written: Generative AI, Functional Generative Networks, FGN, FGNs, Music generation model, Consistency Models, Flash models

AI models that create new content, like text, images, or code, based on patterns learned from training data.

In plain terms

Generative models are a type of artificial intelligence that can create new content. This content can be text, like stories or poems, images, music, or even computer code. They learn by studying vast amounts of existing data and then use that knowledge to produce something original.

For example, you can ask a generative model to write a short story about a robot who discovers music, and it will create a new narrative based on the patterns and styles it learned from reading many stories. Similarly, it can generate an image of a "cat wearing a hat" based on descriptions you provide.

The expert version

Generative models are AI systems designed to produce novel data instances that resemble their training data distribution. They learn a probabilistic representation of the input data, allowing them to sample from this learned distribution to generate new outputs. Common architectures include Generative Adversarial Networks (GANs), Variational Autoencoders (VCEs), and more recently, diffusion models and transformer-based autoregressive models.

These models excel at tasks requiring creativity and synthesis, such as content creation, data augmentation, and simulation. Tradeoffs often involve computational cost for training and inference, potential for generating biased or nonsensical outputs, and challenges in controlling the exact nature of the generated content.

Why it matters now

Generative models are central to recent advancements in AI, enabling frontier labs to launch new products for content creation. Companies like Meta AI are using these models for sophisticated image and video generation, while Anthropic leverages them for specialized applications like financial analysis.

Where it shows up 9 items, newest first
How each company uses it
Anthropic Claude
Anthropic uses generative models, like Claude, for various applications including text generation and specialized analysis.
Meta Meta AI
Meta AI develops generative models for creating new content, such as their Muse Image and Muse Video models.
Related terms

Large Language ModelsDiffusion ModelsGANsAI ArtText GenerationImage Generation

Often confused with
Discriminative models
These models predict a label or category for a given inpu
AI Agents
These are systems that can perceive their environment, make decisions
Foundation Models
These are large models trained on broad data that can be adap
Origin

The concept of generative modeling has roots in statistical modeling and machine learning, with significant advancements driven by deep learning techniques. Early influential work includes research on Generative Adversarial Networks (GANs) introduced by Ian Goodfellow and colleagues in 2014. Source ↗

Try this about 5 minutes

Ask a generative AI chatbot like ChatGPT or Claude to write a short poem about a specific topic, such as 'a rainy day in a city'. Then, try asking it to generate an image based on a simple description like 'a futuristic cityscape at sunset'.

Try it: Playbook entries that use thisPlaybook →

The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.

Questions people ask

What is the difference between generative AI and other AI?
Generative AI creates new content, while other AI types might focus on classifying data, making predictions, or recognizing patterns without producing novel outputs.
Are generative models always accurate?
Generative models can sometimes produce incorrect, nonsensical, or biased information because they learn from existing data, which may contain errors or biases.

First seen on Subvolts Jun 20, 2024. Updated Oct 12, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.