The moment the AI actually 'thinks' and produces an output after you hit enter.
There are two lives of an AI: Training (learning) and Inference (doing). Inference is the act of using the trained model to make a prediction or generate text. Every time you use ChatGPT, you are running inference.
Inference costs are the main expense for running AI companies. Optimizing inference is the battleground for making AI profitable and accessible.
A few adjacent definitions to lock in the concept.
A subset of machine learning that uses multi-layered neural networks to learn from vast amounts of data.
Tags: MechanicsThe process of training a pre-existing AI model on specific data to make it an expert.
Tags: MechanicsA computer system designed to mimic the way the human brain learns and processes information.
Tags: Mechanics