The evolution of ideas

From algorithms
to agents.

First we taught machines to predict. Then to generate. Now we’re teaching systems to act.

A conceptual learning map, not a strict historical timeline. These fields overlap and coexist. Newer approaches do not make earlier ones obsolete.

Statistics
Reason about evidence and uncertainty.
Explore the field →
Machine Learning
Fit patterns and test generalization.
Explore the field →
Deep Learning
Learn layered representations.
Explore the field →
Language & Vision
Represent text, images, and sequences.
Explore the field →
Transformers
Mix information through attention.
Explore the field →
Foundation Models
Adapt pretrained representations and capabilities.
Explore the field →
Generative AI
Generate text and other outputs.
Explore the field →
Retrieval-Augmented Generation
Connect generation to selected external evidence.
Explore the field →
Agents
Connect model decisions to tools and observations.
Explore the field →
Multi-Agent Systems
Coordinate specialized processes when the task justifies it.
Explore the field →
Autonomous AI Systems
Define scope, permissions, monitoring, and stopping conditions.
Explore the field →

From algorithmsto agents.