# AI Literacy The meta-skill of understanding [[Artificial Intelligence (AI)]] well enough to use it effectively. Not programming AI; understanding what it can and cannot do, how it works at a high level, how to evaluate its output, when to trust or distrust it, and how to communicate with it. AI Literacy is the foundation for everything else in the AI-augmented workflow. Without it, people either over-rely on AI (accepting hallucinated output as truth) or under-use it (dismissing it as unreliable). Both failure modes stem from the same root: not understanding the tool. This is different from [[Prompt Engineering]], which is a specific technique within AI Literacy. Prompt Engineering is about crafting effective inputs. AI Literacy is the broader understanding that makes prompt engineering possible and tells you when prompting is even the right approach. Key dimensions of AI Literacy: - **Capability awareness**: knowing the [[Levels of AI use]] and what current models can realistically do - **Failure mode recognition**: understanding [[AI Hallucination]] patterns and [[AI Bias]] so you can spot them - **Output evaluation**: developing judgment about when AI output is trustworthy vs. when it needs verification - **[[Cognitive debt]]** awareness: recognizing when AI use is creating understanding gaps rather than filling them - **Communication skill**: learning to structure requests, provide context, and iterate effectively AI Literacy is not a fixed skill. It evolves as models evolve. What was true about GPT-3 is not true about current frontier models. Staying literate requires ongoing engagement. ## References ## Related - [[Artificial Intelligence (AI)]] - [[Levels of AI use]] - [[AI Hallucination]] - [[AI Bias]] - [[Cognitive debt]] - [[Prompt Engineering]]