TechCrunch publishes a plain-English glossary of common AI terms

In short: TechCrunch published a regularly updated glossary to help people understand common AI terms and newer phrases.

What happened

TechCrunch’s AI team published a glossary that explains AI words and phrases in plain English. The goal is to help readers follow AI news and product updates without needing a technical background.

The glossary covers many commonly used terms, including “large language model” or LLM (the type of AI behind chatbots like ChatGPT), “AI agent” (a tool that can take multiple steps for you, like booking tickets), and “hallucination” (when an AI makes something up that is wrong).

It also adds newer and more debated terms. One example is “opaque recurrence,” described as a way some AI models can “loop” a question through their internal systems instead of showing step-by-step reasoning in readable text. TechCrunch notes that this can make models more efficient, but it can also make it harder for researchers to check what the AI did and why.

Why it matters

AI tools are showing up in more workplaces and everyday apps, and the language around them is getting harder to follow. A glossary like this is useful because it turns confusing labels into simple ideas, like calling API endpoints “buttons” that other software can press. It can also help people spot risks, such as hallucinations, before trusting an AI answer in situations like health or money.

Source: TechCrunch AI