TechCrunch publishes glossary of AI terminology as field's vocabulary expands rapidly
A new resource aims to clarify increasingly technical and fast-moving AI jargon for builders, investors, and observers.
What happened
TechCrunch published a living glossary defining key AI terms including AGI, AI agents, chain-of-thought reasoning, coding agents, compute, deep learning, and others. The publication notes that AI safety researchers have become concerned about techniques like "opaque recurrence" in OpenAI's new Astra model. TechCrunch frames the glossary as a response to the rapid pace of terminology creation in AI, which has made it difficult for even knowledgeable people in tech to keep up with the vocabulary.
Context
The AI field is generating specialized language faster than standard technical communication can absorb it. Multiple definitions exist for core concepts like AGI, with OpenAI, Google DeepMind, and other organizations offering differing frameworks. As AI capabilities expand—particularly in domains like autonomous agents that can operate across multiple systems and interact with third-party services—the precision of terminology becomes increasingly important for builders, investors, and policymakers trying to understand capabilities, risks, and implications. The glossary serves as a reference point in a field where terminology itself is contested and evolving.
What's disputed
The definition of AGI itself is disputed among AI researchers and organizations; sources note that even experts at the forefront of AI research are confused about what AGI actually means. The term "AI agent" is also noted as potentially meaning different things to different people, with infrastructure still being built out to match envisaged capabilities.