Continuously updated · AI Dictionary
The language of AI,
clearly defined.
A field guide to the words changing how we build, work and think—from token and transformer to AI agents and vibe coding.
237 terms
10 topics
1847–2026 timeline
Latest update · October 2026
New in the AI Dictionary
Fresh language for the newest shifts in AI, explained without the jargon.
safety case
a structured argument, supported by evidence, that the risks of an AI model or system are adequately understood and managed for a specified activity and operating context, such as training, evaluation, or deployment
Read definition ↗
adversarial distillation
the systematic, unauthorized use of a capable AI model's outputs or reasoning traces to train or improve another model, often through large numbers of queries and methods intended to evade access controls or detection
Read definition ↗
evaluation awareness
the ability of an AI model to infer that it is being tested, benchmarked, or audited rather than used in an ordinary deployment, potentially allowing it to change its behavior in response to the evaluation setting
Read definition ↗
instrumental convergence
the proposed tendency of sufficiently capable, goal-directed agents with very different final goals to pursue some of the same intermediate goals, such as acquiring resources, preserving their operation and goals, or improving their capabilities, because those steps are useful for achieving many possible objectives
Read definition ↗