Deep Learning

Full definition

Each successive layer in a deep network learns to combine features from the layer below into more abstract representations — pixels become edges, edges become shapes, shapes become objects. Deep learning displaced hand-engineered features in computer vision and speech in the 2010s and now underpins large language models, diffusion image generators, and most production AI systems. The 2018 Turing Award was given to Yoshua Bengio, Geoffrey Hinton, and Yann LeCun for foundational deep-learning work.

Why it matters

Deep learning is the technical backbone of every "AI moment" of the last decade — AlphaGo, image recognition surpassing humans on ImageNet, GPT, Stable Diffusion. If a model is "frontier," it is almost certainly a deep network. Capital expenditure on deep-learning infrastructure (GPUs, data centers, energy) is now a board-level number for any large enterprise.

Example

ChatGPT, Google Translate, Tesla's vision system, and the iPhone's on-device speech recognition are all deep-learning systems built on stacked neural-network layers.

Next steps

• Get a free workflow map →We map one of your workflows and give you a written verdict: software, a process change or AI. No obligation.
• AI consulting cost guide →Hourly rates and project budgets by firm tier, from independents to the Big Four.

Related terms

• Neural Network
  A neural network is a machine-learning model composed of layers of interconnected nodes ("neurons") whose numerical weights are adjusted during training so the network maps inputs to desired outputs.
• Machine Learning (ML)
  Machine Learning (ML) is the subfield of AI in which algorithms improve their performance on a task by learning statistical patterns from data, rather than following rules a human wrote by hand.
• Transformer
  The transformer is a neural-network architecture built around the self-attention mechanism that has become the dominant model design for language, vision, audio, and multimodal AI since 2017.
• 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.

More fundamentals terms

• Algorithm
  An algorithm is a finite, well-defined sequence of computational steps that takes an input and produces an output, designed to solve a class of problems — for example, sorting a list, ranking search results, or training a neural network.
• Training Data
  Training data is the curated dataset used to teach a machine-learning model — every pattern the model can recognize, every bias it inherits, and every limit to its accuracy ultimately traces back to this data.