Continual Learning

Using AI

The ability of an AI system to keep learning new things over time without forgetting what it already knows.

Most models are frozen after training - they learn nothing from use, and retraining on new data risks catastrophic forgetting, where new knowledge overwrites old skills. Continual learning is the research problem of updating models safely and incrementally, the way people learn.It matters for systems that live in changing environments: assistants that should absorb new products and policies, robots facing new situations, fraud models chasing evolving scams. Today most products fake it by retraining periodically or bolting on retrieval rather than truly learning continuously.