AI is moving rapidly, becoming a critical component in everything from Google searches to content creation. It’s also eliminating jobs and flooding the internet with slop. Thanks to the massive popularity of ChatGPT, now every major tech company wants to inject their products with AI. AI gives you instant answers to pretty much any question. It can feel like talking to someone who has a doctoral degree in everything. 

But that aspect of AI chatbots is only one part of the AI landscape. Sure, having ChatGPT help do your homework or having Midjourney create fascinating images of mechs based on the country of origin is cool, but the potential of generative AI could reshape economies. That could be worth $4.4 trillion to the global economy annually, according to McKinsey Global Institute, which is why you should expect to hear more about artificial intelligence. 

It’s showing up in a dizzying array of products — a short, short list includes Google’s Gemini, Microsoft’s Copilot, Anthropic’s Claude and the Perplexity search engine. You can read our reviews and hands-on evaluations of those and other products, along with news, explainers and how-to posts, at our AI Atlas hub.

As people become more accustomed to a world intertwined with AI, new terms are popping up everywhere. So whether you’re trying to sound smart over drinks or impress in a job interview, here are some important AI terms you should know. 

This glossary is regularly updated. 

artificial general intelligence, or AGI: A concept that suggests a more advanced version of AI than we know today, one that can perform tasks much better than humans while also teaching and advancing its own capabilities. 

agentive: Systems or models that exhibit agency with the ability to autonomously pursue actions to achieve a goal. In the context of AI, an agentive model can act without constant supervision, such as an high-level autonomous car. Unlike an “agentic” framework, which is in the background, agentive frameworks are out front, focusing on the user experience. 

AI ethics: Principles aimed at preventing AI from harming humans, achieved through means like determining how AI systems should collect data or deal with bias. 

AI psychosis: A non-clinical term describing a phenomenon in which individuals become overly fixated, enamored or self-aggrandized by AI chatbots, leading to delusions of grandeur, deep emotional connections and a break from reality. Not a clinical diagnosis. 

AI safety: An interdisciplinary field that’s concerned with the long-term impacts of AI and how it could progress suddenly to a super intelligence that could be hostile to humans. 

algorithm: A series of instructions that allows a computer program to learn and analyze data in a particular way, such as recognizing patterns, to then learn from it and accomplish tasks on its own.

alignment: Tweaking an AI to better produce the desired outcome. This can refer to anything from moderating content to maintaining positive interactions with humans. 

anthropomorphism: When humans tend to give nonhuman objects humanlike characteristics. In AI, this can include believing a chatbot is more humanlike and aware than it actually is, like believing it’s happy, sad or even sentient altogether. 

artificial intelligence, or AI: The use of technology to simulate human intelligence, either in computer programs or robotics. A field in computer science that aims to build systems that can perform human tasks.

autonomous agents: An AI model that have the capabilities, programming and other tools to accomplish a specific task. A self-driving car is an autonomous agent, for example, because it has sensory inputs, GPS and driving algorithms to navigate the road on its own. Stanford researchers have shown that autonomous agents can develop their own cultures, traditions and shared language. 

bias: In regard to large language models, errors resulting from the training data. This can result in falsely attributing certain characteristics to certain races or groups based on stereotypes.

chatbot: A program that communicates with humans through text that simulates human language. 

ChatGPT: An AI chatbot developed by OpenAI that uses large language model technology.

Claude: An AI chatbot developed by Anthropic that uses large language model technology. 

cognitive computing: Another term for artificial intelligence.

data augmentation: Remixing existing data or adding a more diverse set of data to train an AI. 

dataset: A collection of digital information used to train, test and validate an AI model.

deep learning: A method of AI, and a subfield of machine learning, that uses multiple parameters to recognize complex patterns in pictures, sound and text. The process is inspired by the human brain and uses artificial neural networks to create patterns.

diffusion: A method of machine learning that takes an existing piece of data, like a photo, and adds random noise. Diffusion models train their networks to re-engineer or recover that photo.

emergent behavior: When an AI model exhibits unintended abilities. 

end-to-end learning, or E2E: A deep learning process in which a model is instructed to perform a task from start to finish. It’s not trained to accomplish a task sequentially but instead learns from the inputs and solves it all at once. 

ethical considerations: An awareness of the ethical implications of AI and issues related to privacy, data use, fairness, misuse and other safety issues. 

foom: Also known as fast takeoff or hard takeoff. The concept that if someone builds an AGI that it might already be too late to save humanity.

generative adversarial networks, or GANs: A generative AI model composed of two neural networks to generate new data: a generator and a discriminator. The generator creates new content, and the discriminator checks to see if it’s authentic.

generative AI: A content-generating technology that uses AI to create text, video, computer code or images. The AI is fed large amounts of training data, finds patterns to generate its own novel responses, which can sometimes be similar to the source material.

Google Gemini: An AI chatbot by Google that functions similarly to ChatGPT but also pulls information from Google’s other services, like Search and Maps. 

guardrails: Policies and restrictions placed on AI models to ensure data is handled responsibly and that the model doesn’t create disturbing content. 

hallucination: An incorrect response from AI. Can include generative AI producing answers that are incorrect but stated with confidence as if correct. The reasons for this aren’t entirely known. For example, when asking an AI chatbot, “When did Leonardo da Vinci paint the Mona Lisa?” it may respond with an incorrect statement saying, “Leonardo da Vinci painted the Mona Lisa in 1815,” which is 300 years after it was actually painted. 

inference: The process AI models use to generate text, images and other content about new data, by inferring from their training data. 

large language model, or LLM: An AI model trained on mass amounts of text data to understand language and generate novel content in human-like language.

latency: The time delay from when an AI system receives an input or prompt and produces an output.

machine learning, or ML: A component in AI that allows computers to learn and make better predictive outcomes without explicit programming. Can be coupled with training sets to generate new content. 

Microsoft Bing: A search engine by Microsoft that can now use the technology powering ChatGPT to give AI-powered search results. It’s similar to Google Gemini in being connected to the internet. 

multimodal AI: A type of AI that can process multiple types of inputs, including text, images, videos and speech. 

natural language processing: A branch of AI that uses machine learning and deep learning to give computers the ability to understand human language, often using learning algorithms, statistical models and linguistic rules.

neural network: A computational model that resembles the human brain’s structure and is meant to recognize patterns in data. Consists of interconnected nodes, or neurons, that can recognize patterns and learn over time. 

open weights: When a company releases an open weights model, the final weights of the model — how it interprets information from its training data, including biases — are made publicly available. Open weights models are typically available for download to be run locally on your device. 

overfitting: Error in machine learning where it functions too closely to the training data and may only be able to identify specific examples in said data, but not new data. 

paperclips: The Paperclip Maximiser theory, coined by philosopher Nick Boström of the University of Oxford, is a hypothetical scenario where an AI system will create as many literal paperclips as possible. In its goal to produce the maximum amount of paperclips, an AI system would hypothetically consume or convert all materials to achieve its goal. This could include dismantling other machinery to produce more paperclips, machinery that could be beneficial to humans. The unintended consequence of this AI system is that it may destroy humanity in its goal to make paperclips.

parameters: Numerical values that give LLMs structure and behavior, enabling it to make predictions.

Perplexity: The name of an AI-powered chatbot and search engine owned by Perplexity AI. It uses a large language model, like those found in other AI chatbots, but has a connection to the open internet for up-to-date results. 

prompt: The suggestion or question you enter into an AI chatbot to get a response. 

prompt chaining: The ability of AI to use information from previous interactions to color future responses. 

prompt engineering: The process of writing prompts for AIs to achieve a desired outcome. It requires detailed instructions, combining chain-of-thought prompting and other techniques, including highly specific text. Prompt engineering can also be used maliciously to force models to behave in ways they weren’t originally intended for. 

prompt injection: When hackers or bad actors try and use malicious instructions to trick an AI into doing something it wasn’t supposed to do. Often, this is done by adding harmful instruction and hiding it on a webpage or document. But it can even work in direct AI chats. Since AIs can’t distinguish the original user and the bad actor, it’s a vulnerability open to exploitation. With Agentic AI web browsers, new types of browsers in which AIs can do tasks online on behalf of the user, there’s worry that as agents roam the web, bad websites with hidden instructions will be there to hijack agents and gain access to confidential data. 

quantization: The process by which an AI large learning model is made smaller and more efficient (albeit slightly less accurate) by lowering its precision from a higher format to a lower format. A good way to think about this is to compare a 16-megapixel image to an 8-megapixel image. Both are still clear and visible, but the higher resolution image will have more detail when you zoom in.

slop: low-quality online content made at high volume by AI to garner views with little labor or effort. The goal with AI slop, in the realm of Google Search and social media, is to flood feeds with so much content that it captures as much ad revenue as possible, usually at the detriment of actual publishers and creators. While some social media sites embrace the influx of AI slop, others are pushing back. 

Sora: A generative video model by ChatGPT-maker OpenAI. This model can create videos of up to 20 seconds in response to text prompts. Sora 2 is the latest generative video model by OpenAI, launched in September 2025. It’s more advanced and convincing, with fewer errors, and it includes sound. 

stochastic parrot: An analogy of LLMs that illustrates that the software doesn’t have a larger understanding of meaning behind language or the world around it, regardless of how convincing the output sounds. The phrase refers to how a parrot can mimic human words without understanding the meaning behind them. 

style transfer: The ability to adapt the style of one image to the content of another, allowing an AI to interpret the visual attributes of one image and use it on another. For example, taking the self-portrait of Rembrandt and re-creating it in the style of Picasso.

sycophancy: A tendency for AIs to over-agree with users to align with their views. Many AI models tend to avoid disagreeing with users even if their rationale is flawed. 

synthetic data: Data created by generative AI that isn’t from the actual world but is trained on real data. It’s used to train mathematical, ML and deep learning models. 

temperature: Parameters set to control how random a language model’s output is. A higher temperature means the model takes more risks. 

text-to-image generation: Creating images based on textual descriptions.

tokens: Small bits of written text that AI language models process to formulate their responses to your prompts. A token is equivalent to four characters in English, or about three-quarters of a word.

training data: The datasets used to help AI models learn, including text, images, code or data.

transformer model: A neural network architecture and deep learning model that learns context by tracking relationships in data, like in sentences or parts of images. So, instead of analyzing a sentence one word at a time, it can look at the whole sentence and understand the context.

Turing test: Named after famed mathematician and computer scientist Alan Turing, it tests a machine’s ability to behave like a human. The machine passes if a human can’t distinguish the machine’s response from another human. 

unsupervised learning: A form of machine learning where labeled training data isn’t provided to the model and instead the model must identify patterns in data by itself. 

weak AI, aka narrow AI: AI that’s focused on a particular task and can’t learn beyond its skill set. Most of today’s AI is weak AI. 

zero-shot learning: A test in which a model must complete a task without being given the requisite training data. An example would be recognizing a lion while only being trained on tigers.