{
“title”: “AI Glossary: Master Today’s Essential Tech Terms”,
“Text1”: “
AI isn’t just changing how we work. It’s inventing a whole new language to describe itself, and that vocabulary moves fast. TechCrunch AI published a living glossary of the terms you’re most likely to hear in a pitch, a product meeting, or a panel, and it’s worth knowing cold. As TechCrunch AI notes, even very smart people in tech feel a little insecure when the jargon starts flying, including newer terms like “opaque recurrence,” the reasoning technique in OpenAI’s new Astra model that has safety researchers rattled.
\n\n
What stands out here is how many of these terms get used loosely, even by experts. This guide walks through each one in order, with plain-English definitions and a note on why each matters.
\n\n
Quick Start: what you’ll learn
\n\n
By the end, you’ll be able to define the core AI terms floating around today: AGI, AI agents, API endpoints, chain-of-thought reasoning, coding agents, compute, and deep learning. No prior knowledge needed. Just read top to bottom.
\n\n
AGI (Artificial General Intelligence)
\n\n
A nebulous term for AI that’s more capable than the average human at many or most tasks. Definitions vary by lab. Sam Altman once called it the “equivalent of a median human that you could hire as a co-worker.” OpenAI’s charter frames it as “highly autonomous systems that outperform humans at most economically valuable work.” Google DeepMind says AI “at least as capable as humans at most cognitive tasks.”
\n\n
Why it matters: if the experts can’t agree, treat any confident AGI claim with skepticism.
\n\n
AI agent
\n\n
A tool that uses AI to perform a series of tasks for you, going beyond a basic chatbot. Think filing expenses, booking a table, or writing and maintaining code. It’s an autonomous system that may draw on multiple AI systems to handle multistep tasks.
\n\n
Why it matters: the term means different things to different people, and the infrastructure is still being built. Ask what someone actually means when they say “agent.”
\n\n
API endpoints
\n\n
Think of these as hidden “buttons” on the back of software that other programs can press to make it do things. Developers use them to build integrations, like letting one app pull data from another. Most smart home devices have them, even if you never see them.
\n\n
Why it matters: as AI agents get more capable, they can find and use these endpoints on their own, which opens up powerful and sometimes unexpected automation.
\n\n
Chain-of-thought reasoning
\n\n
Breaking a problem into smaller, intermediate steps to improve the final answer. Some questions are instant (“which is taller, a giraffe or a cat?”). Others need pen and paper, like figuring out that 40 heads and 120 legs means 20 chickens and 20 cows.
\n\n
Why it matters: this approach takes longer but is more likely to be correct, especially in logic or coding. Reasoning models are built from standard large language models and optimized for this through reinforcement learning.
\n\n
Coding agent
\n\n
A specialized agent for software development. Rather than just suggesting code for a human to paste in, it can write, test, and debug autonomously, working across entire codebases and pushing fixes with minimal oversight. TechCrunch AI compares it to hiring a very fast intern who never sleeps. The catch, as with any intern, is that a human still needs to review the work.
\n\n
Compute
\n\n
Shorthand for the computational power that lets AI models run. It fuels the whole industry, from training to deployment, and often refers to the hardware behind it: GPUs, CPUs, TPUs, and other infrastructure.
\n\n
Why it matters: compute is the bedrock, and access to it shapes who can build at the frontier.
\n\n
Deep learning
\n\n
A subset of machine learning built on multi-layered artificial neural networks, loosely inspired by the neurons in the human brain. These models find important patterns in data on their own instead of waiting for engineers to define them, and they improve through repetition. The trade-off: they need millions of data points and take longer to train, so costs run higher.
\n\n
Your next steps
\n\n
Keep the glossary bookmarked. TechCrunch AI updates it as the field evolves, so treat it as a living document. Next time you hear a term you don’t know, don’t nod along. Look it up, and pay attention to how loosely people use words like “agent” or “AGI.” You can find the full and continually updated list at the original source.
”
}