source: techcrunch ai: the only ai glossary you’ll need this year
level: business
artificial intelligence is creating a new vocabulary. terms like llm, rag, and rlhf appear in meetings and articles. this glossary defines key ai concepts in simple language. it covers agi, ai agents, api endpoints, chain of thought, coding agents, compute, deep learning, diffusion, distillation, fine-tuning, gans, hallucinations, inference, large language models, memory cache, model context protocol, mixture of experts, and neural networks. the list is updated regularly as the field changes.
agi is ai that matches or exceeds human ability across many tasks. definitions vary: openai sees it as a median human coworker, while google deepmind focuses on cognitive tasks. ai agents perform multistep tasks like booking tickets or writing code, going beyond simple chatbots. api endpoints let software interact, and ai agents can use them for automation. chain of thought breaks problems into steps for better reasoning, especially in logic or coding. coding agents autonomously write, test, and debug code. compute refers to the processing power from gpus and other hardware that trains and runs ai models.
deep learning uses multi-layered neural networks to find patterns in data without human-defined features. it needs lots of data and compute. diffusion models create images or music by learning to reverse a noising process. distillation trains a smaller student model from a larger teacher model, often to improve efficiency. fine-tuning adapts a pre-trained model for specific tasks with new data. gans pit two networks against each other to generate realistic outputs. hallucinations are incorrect ai-generated information, a major quality issue. inference is running a trained model to make predictions. llms are large neural networks powering chatbots like chatgpt. memory cache speeds up inference by storing calculations. mcp is an open standard for connecting ai to external tools. mixture of experts uses specialized subnetworks to handle tasks efficiently.
why it matters: understanding these terms helps professionals evaluate ai tools, communicate with teams, and make informed decisions about adopting or investing in ai technology.
source: techcrunch ai: the only ai glossary you’ll need this year