In this lesson you’ll cover:
- A working definition of an “agent” and a close reading of what each part of that definition means (autonomy, goal pursuit, operating without per-step human direction)
- The five traits that define whether a system qualifies as an agent: autonomy, goal-directed behavior, tool use, state persistence, and adaptability
- A taxonomy of what counts as an agent versus what doesn’t, with examples of deployed agentic systems
- The REAC framework (Retrieve, Evaluate, Act, Combine) as a functional description of how agents operate step by step
- How agents formulate multiple sub-queries, evaluate source credibility, adapt mid-process, and synthesize a final answer
- The synthesis methods agents use: claim aggregation, complementary information, conflict resolution, gap filling, and attribution
- A point-by-point contrast between traditional crawlers and agents across purpose, path selection, evaluation timing, and multi-source usage
- Where SEO work actually connects to agent selection