Term

Machine Learning (ML)

Machine learning is the field that builds systems which improve at a task by learning patterns from data rather than from explicit hand-coded rules. The field has three classical families: supervised learning (the model learns from labeled examples), unsupervised learning (the model finds structure in unlabeled data), and reinforcement learning (the model learns from environmental reward). Generative AI and LLMs sit primarily in the supervised and self-supervised lineage.
Reviewed by 7wData
Mentioned Large Language Model (LLM)

Why it matters

The AI conversation in 2026 over-indexes on generative AI and LLMs, but the bulk of production ML inside enterprises is still classical supervised models doing credit scoring, fraud detection, demand forecasting, and churn prediction. I am seeing risk teams reinvent practices that already exist. Model risk management (validation, drift monitoring, lineage tracking) was built around these classical models long before LLMs arrived. The frameworks extend cleanly to LLMs once teams remember the foundations. Treating LLMs as a separate species from the rest of ML is how you end up with two parallel governance stacks that contradict each other.

Where you’ll encounter it

Three contexts. First, a vendor pitches “AI” but the substrate under the demo is a gradient-boosted tree (classical ML, mature, well-understood, and often the right answer). Second, a model risk review references SR 11-7, the US Federal Reserve guidance on model risk management, which predates the LLM wave and is anchored in classical ML assumptions about validation and monitoring. Third, a data team’s job description distinguishes “ML Engineer” from “Data Scientist”, and the boundary is fuzzy and depends on the org.

Part of the 7wData AI Glossary. Tracking how concepts like this move in the expert conversation: daily signals at ins7ghts.com.