TechCrunch has published a glossary of AI terms to help readers navigate the rapidly evolving vocabulary of artificial intelligence, including opaque recurrence — the reasoning technique in OpenAI's new Astra model that has raised concerns among AI safety researchers. The guide aims to provide plain-English definitions for terms like LLMs, RAG, and RLHF that even tech-savvy individuals may find confusing.
The glossary covers several key concepts, including artificial general intelligence (AGI), which has varying definitions depending on the source. OpenAI CEO Sam Altman has described AGI as "equivalent of a median human that you could hire as a co-worker," while OpenAI's charter defines it as "highly autonomous systems that outperform humans at most economically valuable work." Google DeepMind offers a slightly different view, defining AGI as "AI that's at least as capable as humans at most cognitive tasks."
Additional terms explained include AI agents, which are tools that use AI to perform multistep tasks beyond basic chatbot capabilities — such as filing expenses, booking reservations, or writing and maintaining code. API endpoints are described as "buttons" on software that allow other programs to interact with them, enabling integrations and automation. The glossary is positioned as a living document that will be regularly updated as the field evolves.