Also known as: artificial neural network, ANN
In plain English
A neural network is loosely inspired by the brain. It's made of many small connected units that pass numbers to each other. During training, the connections are adjusted again and again until the network gives good answers.
In practice
Neural networks are the building blocks of nearly all modern AI models. Their weakness is explainability: it's hard to say exactly why a network made a particular decision, which matters for regulated decisions like lending.
Under the hood
A neural network composes layers of linear transformations followed by non-linear activation functions. Its weights are learned by backpropagating the gradient of a loss function. Model size is usually described by its parameter count.
Example
"The neural network has seven billion parameters."