The only AI glossary you’ll need this year

A new glossary aims to demystify the rapidly expanding lexicon of artificial intelligence, offering plain-English definitions for terms from AGI to deep learning that have become common in product meetings and investment pitches [1]. The guide, published by TechCrunch, defines artificial general intelligence (AGI) as AI that is more capable than the average human at many, if not most, tasks [1]. OpenAI CEO Sam Altman has described AGI as "the equivalent of a median human that you could hire as a co-worker" [1]. The lab Google DeepMind, founded in the UK in 2010 and acquired by Google in 2014, views AGI as "AI that's at least as capable as humans at most cognitive tasks" [1][8]. The field of AI research was formally founded at a workshop at Dartmouth College in 1956, where attendees predicted machines as intelligent as humans would exist within a generation [3]. AI agents are tools that use AI technologies to perform tasks on a user's behalf, such as filing expenses or writing and maintaining code [1]. A 2025 survey of 25 leading researchers from labs including Google DeepMind, OpenAI, and Anthropic found that 20 identified automating AI research itself as one of the most severe and urgent AI risks [5]. Participants predicted AI agents will gradually transition from assistants to autonomous AI developers, though they disagreed on timelines and governance mechanisms [5]. The glossary explains that API endpoints function as "buttons" on the back of software that other programs can press to make it do things [1]. Chain-of-thought reasoning, meanwhile, involves breaking a problem into smaller intermediate steps to improve the quality of the end result [1]. The guide illustrates this with an example: if a farmer has chickens and cows, and together they have 40 heads and 120 legs, the answer is 20 chickens and 20 cows [1]. Coding agents are specialized AI agents that can write, test, and debug code autonomously [1]. Deep learning, a subset of machine learning, uses multi-layered artificial neural network structures that allow algorithms to make complex correlations and improve their own outputs through repetition and adjustment [1]. These systems require millions of data points to yield good results [1]. The transformer architecture, introduced in 2017, was utilized to produce generative AI applications and led to the rapid scaling of large language models like ChatGPT [3]. Generative AI, which includes chatbots such as ChatGPT, Claude, and Google Gemini, has seen a significant increase in prevalence since the AI boom of the 2020s [2]. Companies in sectors including software development, healthcare, and finance have adopted the technology [2]. However, concerns persist. Algorithmic bias describes systematic and repeatable harmful tendencies in computerized systems that can reinforce social biases of race, gender, and ethnicity [4]. The European Union's Artificial Intelligence Act, proposed in 2021 and adopted in 2024, is among the legal frameworks beginning to address such discrimination [4].