TerminologytermStep 1: Start hereAllRiskHR

Term: Machine learning (ML)

~6 min read

Estimated time: ~6 min read — for the in-app brief plus opening the primary source.

What this is

Machine learning is how most modern AI gets its skill: it learns patterns from examples instead of following only hand-written rules.

Everyday example

A spam filter that improves as it sees more mail, or a hiring screen that learned from past résumés, is machine learning — skill from examples, not from a rule someone typed.

Machine learning is how most modern AI gets its skill: it learns patterns from examples, not from hand-written rules.

  • A rules engine says “if amount > $10k, flag.” A learned model infers patterns from past cases.
  • Quality follows the data: biased or stale examples become biased or stale decisions.
  • Generative models are a form of machine learning at very large scale.
  • You do not need the math — you do need to ask what it learned from, and how it is re-tested.

Next action: For any “smart” tool, ask what examples it learned from and when that set was last refreshed.

What changes in how you lead

How decision rights, process, and ownership should change.

  • Historical data is not a neutral teacher — it encodes past process and past bias.
  • HR, Risk, and Legal should see the training-data story for people and credit decisions.

Compare related ideas

Machine learning (ML) vs Generative AI

Machine learning is the method. Generative AI is one use of that method to create content. Many ML systems never write a sentence.

Open Generative AI

Deep dive

A rules engine says “if amount > $10k, flag.” A learned model infers patterns from history.

Generative models are machine learning at very large scale.

Ask what it learned from, when that set was refreshed, and how it is re-tested.

HR and credit tools are where this becomes a legal conversation, not only a technical one.

Related terms

Related weekly lessons

terminologymldata