AI Categories: Generalisation abilities

Generalisation Ability
Generalisation ability describes how an Artificial General Intelligence (AGI) can take what it has learned in one situation and apply it to completely new, unseen contexts. Unlike narrow AI, which is usually restricted to a specific function, AGI can carry knowledge across different areas—for instance, using ideas gained from studying physics to help design machinery or diagnose a software problem.

This skill is considered a key indicator of true intelligence because it demonstrates adaptability, abstract thinking, and flexible reasoning. Instead of relying on memorised answers, an AGI with strong generalisation can recognise deeper patterns and use them in creative or unfamiliar scenarios. This makes it capable of functioning well in real-world environments where challenges rarely follow a predictable structure.

Generalisation is what enables AGI to move beyond repetitive task-based performance and show genuine problem-solving breadth, similar to how humans use varied past experiences to tackle new situations.