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

Neural network

Also written: ANN, MLP neurons, Artificial Neuron, Neural activity, Graph Neural Net, GNN, Recurrent Neural Networks, RNNs

A system of interconnected units, like a digital brain, that AI uses to learn from data and perform tasks.

In plain terms

A neural network is a core part of many AI systems. Think of it as a computer program designed to learn from information, much like a human brain learns from experience.

It's made up of many small, interconnected processing units, often called nodes or neurons. These nodes work together, passing signals to each other. By adjusting the strength of these connections based on the data it sees, the neural network learns to recognize patterns, make predictions, or generate new content.

For example, when you upload a photo to an AI that identifies objects, a neural network is likely working behind the scenes. It has been trained on millions of images to learn what different objects look like, so it can tell you if it sees a cat, a dog, or a car.

The expert version

A neural network is a computational model inspired by the structure and function of biological neural networks. It consists of layers of interconnected nodes (neurons) that process information. Each connection has a weight, and during training, these weights are adjusted to minimize the difference between the network's output and the desired output for a given input.

Different architectures, such as feedforward networks, convolutional neural networks (CNNs), and recurrent neural networks (RNNs), are suited for different tasks. CNNs excel at image processing, while RNNs are used for sequential data like text or time series. The complexity and size of neural networks have grown significantly, leading to models with billions or even trillions of parameters.

Why it matters now

Neural networks are the core technology behind modern AI systems. Recent advancements and research, like Anthropic's Golden Gate Claude demo, show how researchers are gaining finer control over these networks to shape AI behavior. The ongoing development of increasingly large and capable models by companies like OpenAI and Google Gemini relies heavily on sophisticated techniques for training these complex neural networks.

Where it shows up 23 items, newest first
How each company uses it
Anthropic Claude
Anthropic's Golden Gate Claude research demo specifically highlighted tuning neural features for precise control over AI behavior.
Google Gemini
Google's work on systems like AlphaGo and Gemini is built upon advanced neural network architectures, enabling them to tackle complex problems and understand information.
OpenAI ChatGPT
OpenAI has published techniques for training large neural networks, essential for developing powerful models like those used in ChatGPT and DALL·E.
Related terms

Machine learningDeep learningArtificial intelligenceModel trainingParametersNeurons

Often confused with
Artificial intelligence
AI is a broader field that encompasses neural networks,
Deep learning
Deep learning is a subfield of machine learning that uses neural
Origin

The concept of artificial neural networks was first proposed in the 1940s, inspired by early understandings of biological neurons. Key developments in the 1980s and 1990s, including backpropagation, laid the groundwork for modern neural networks. Source ↗

Try this about 5 minutes

Ask an AI chatbot like ChatGPT or Claude to write a short story about a robot learning to paint. Observe how it generates creative text, demonstrating the pattern recognition and generation capabilities of its underlying neural network.

Try it: Playbook entries that use thisPlaybook →

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

Questions people ask

Are neural networks the same as AI?
Neural networks are a type of AI, but not all AI uses neural networks. They are a powerful tool within the broader field of artificial intelligence.
How do neural networks learn?
They learn by processing large amounts of data and adjusting the connections between their internal units to improve accuracy on specific tasks, a process often called training.

First seen on Subvolts Apr 30, 2020. Updated Oct 12, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.