One-Sentence Definition
Deep learning is a branch of machine learning that uses multi-layered neural networks to automatically learn representations from raw data, enabling breakthroughs in vision, language, and audio processing.
How It Works
A deep-learning model is a neural network with many layers stacked on top of each other -- hence the word "deep." Each layer transforms its input into a slightly more abstract representation. In an image classifier, the first layer might detect edges, the next layer combines edges into shapes, and deeper layers recognize objects like faces or cars. The network learns these representations on its own during training; no one manually programs what an "edge" is.
Training works by feeding data through the network, measuring how far its output is from the correct answer (using a loss function), and then propagating that error backward through the layers (backpropagation) to adjust the weights. This cycle repeats millions of times over large datasets.
Deep learning took off around 2012 when a convolutional neural network called AlexNet won the ImageNet competition by a large margin. Three factors converged: GPUs made parallel computation affordable, the internet provided massive training datasets, and algorithmic improvements like dropout and batch normalization kept deep networks from overfitting. Since then, the field has produced transformers (the architecture behind GPT and Claude), diffusion models (behind Midjourney and DALL-E), and other designs that define the current AI landscape.
Why It Matters
Deep learning is the reason AI went from a niche research topic to a mainstream technology. It powers Google Translate, Tesla's self-driving stack, Spotify's recommendation engine, and every large language model on the market. Before deep learning, tasks like accurate speech recognition and real-time object detection were unsolved problems. Now they ship as features in consumer devices.
The tradeoff is cost. Training frontier deep-learning models requires thousands of GPUs running for weeks, which is why the field is increasingly dominated by well-funded labs like OpenAI, Google DeepMind, Anthropic, and Meta AI.
Key Takeaway
Deep learning uses layered neural networks to learn directly from raw data, and it is the core technology behind virtually every major AI advance since 2012.
Part of the AI Weekly Glossary.