AI glossary, plain English.

Adobe is a leading global software company renowned for its suite of creative and productivity applications, like Photoshop and Illustrator. They are increasingly integrating artificial intelligence, such as their Firefly generative AI models, into their products to enhance creative workflows.

Amazon is a global technology company known for its e-commerce platform, cloud computing services (AWS), digital streaming, and artificial intelligence innovations.

AMD, or Advanced Micro Devices, is a major global semiconductor company known for designing and producing computer processors, including CPUs and GPUs, which are crucial components for everything from personal computers to large data centers and AI systems.

Anthropic is an AI safety and research company known for developing large language models like the Claude family, with a focus on creating beneficial and safe AI systems.

AWS, or Amazon Web Services, is a massive collection of cloud computing services from Amazon. It provides on-demand computing power, storage, databases, and a wide array of other tools over the internet, allowing businesses to run their operations without owning their own physical hardware.

Baidu is a major Chinese technology company, best known for its internet search engine and significant investments in artificial intelligence, including autonomous driving and large language models.

DeepMind is an artificial intelligence research laboratory and company, now part of Google, known for its work in reinforcement learning and artificial general intelligence.

DeepMotion is a company that uses AI to convert videos of people into 3D animations, making it easier to create characters that move realistically in digital worlds.

DeepSeek is an artificial intelligence company known for developing a family of powerful large language models, many of which are released as open-source, meaning their inner workings are publicly available.

ElevenLabs is an AI company specializing in high-quality speech synthesis and voice AI technologies, allowing users to generate realistic speech from text and even create custom AI voices.

GitLab is a comprehensive platform for software development, offering a complete set of tools for version control, continuous integration, security, and project management, all within a single application.

Google is one of the world's largest and most influential technology companies, famous for its search engine. It's a major player in artificial intelligence, developing everything from foundational AI research to everyday AI-powered products like Android, YouTube, and its own large language models.

Google Cloud is Google's collection of cloud computing services, providing businesses and developers with a wide array of tools for computing, storage, networking, and artificial intelligence, all hosted on Google's global infrastructure.

HeyGen is an AI company that develops tools for creating generative videos, allowing users to produce realistic videos with AI avatars speaking from text scripts without needing cameras or actors.

HubSpot is a company that creates and sells software products for marketing, sales, and customer service, aiming to help businesses grow.

IBM, or International Business Machines, is a long-standing American technology and consulting company that has played a significant role in the history of computing and continues to be active in artificial intelligence research and development.

Intel is a historic technology company, most famous for designing and manufacturing the central processing units, or 'brains,' found in countless personal computers and servers. They are also a significant developer of hardware and software for artificial intelligence.

JetBrains is a software company famous for creating powerful developer tools, especially Integrated Development Environments (IDEs) like PyCharm and IntelliJ IDEA, which help programmers write, test, and manage their code more efficiently.

Meta Platforms, formerly Facebook, is a major technology company known for its social media platforms, virtual reality efforts, and significant contributions to artificial intelligence research, including popular open-source large language models.

Microsoft is a multinational technology company known for developing computer software, consumer electronics, and services, with major investments and advancements in artificial intelligence.

Microsoft AI refers to the artificial intelligence initiatives, products, and research efforts led by Microsoft Corporation, a major technology company.

GitHub is a web-based platform that helps developers store, manage, and share their code. It's owned by Microsoft and acts like a social network for software projects.

Mistral AI is a French artificial intelligence company that develops open-source language models. They are known for creating powerful yet efficient AI models that can run on various devices.

MIT, or the Massachusetts Institute of Technology, is a world-renowned university famous for its research and education in science, engineering, and technology. It has been a major contributor to advancements in artificial intelligence.

Mozilla is a global community and organization dedicated to fostering an open, accessible, and user-controlled internet. They are best known for developing the Firefox web browser and advocating for online privacy and open web standards.

MTP AI is a company that develops and implements customized artificial intelligence solutions for businesses, often specializing in computer vision and industrial automation to improve efficiency.

NVIDIA is an American technology company founded in 1993, best known for designing graphics processing units (GPUs) which are essential for modern AI.

NVIDIA is a leading technology company that designs graphics processing units (GPUs), which are crucial for AI, and develops platforms for various computing tasks, from gaming to data centers.

OpenAI is an American artificial intelligence research and deployment company known for creating advanced AI models like ChatGPT and DALL-E. Their mission is to ensure that artificial general intelligence (AGI) benefits all of humanity.

Together AI is a cloud platform that helps developers build, run, and fine-tune large language models (LLMs) and other AI models faster and more affordably.

Voiceflow is a company that provides a design and development platform for building conversational AI experiences, such as chatbots and voice assistants. It offers a visual, collaborative workspace to create, prototype, and deploy AI-powered conversations without needing extensive coding.

An agent is an AI system that can perceive its environment, make decisions, and take actions to achieve specific goals, often by breaking down complex tasks into smaller steps.

Artificial General Intelligence, or AGI, refers to hypothetical AI that can understand, learn, and apply intelligence to any intellectual task a human being can.

The context window refers to the amount of information an AI model can 'see' and process at any given time when generating a response. It's like the AI's short-term memory.

A "copilot" is an AI assistant designed to work alongside a human, helping with tasks like writing, coding, or data analysis by offering suggestions and automating parts of the work.

CSP stands for Constraint Satisfaction Problem. It's a way to think about problems where you need to find values for variables that satisfy certain limits or rules.

CX stands for Customer Experience, which is all about how a customer perceives their interactions with a company.

Deep learning is a type of machine learning that uses multi-layered artificial neural networks to learn from data, often without direct human instruction.

DevOps is a set of practices that combines software development (Dev) with IT operations (Ops) to shorten the systems development life cycle and provide continuous delivery.

Embeddings are numerical representations of words, phrases, or other data, capturing their meaning and relationships in a way computers can understand.

Explainable AI, or XAI, is all about making AI systems understandable to humans, so we can see how they make decisions.

MiniMax is a decision-making rule used in game theory and artificial intelligence to minimize the possible loss for a worst-case scenario. It helps an AI choose the best move in two-player games by assuming the opponent plays optimally.

Moonshot AI refers to highly ambitious, often speculative artificial intelligence projects that aim to solve previously intractable problems or achieve breakthroughs with massive societal impact. They often involve significant risk and long development timelines.

Multimodal refers to AI systems that can process, understand, and generate information using more than one type of data, such as combining text with images, audio, or video. It allows AI to interact with the world in a richer, more human-like way.

Natural Language Processing, or NLP, is an area of AI that focuses on enabling computers to understand, interpret, and generate human language.

Open source refers to software or projects where the original source code is made freely available and can be modified and distributed by anyone.

PII stands for Personally Identifiable Information, which is any data that can be used to identify a specific individual. Protecting PII is crucial in the age of digital information.

A reasoner is a system that can draw conclusions or infer new facts from existing knowledge and a set of rules. It helps AI systems understand and make sense of information.

SEO stands for Search Engine Optimization. It's the process of improving your website so it ranks higher in search engine results, making it easier for people to find you online.

SWE stands for Software Engineer, a professional who designs, develops, and maintains software systems.

UMAP, short for Uniform Manifold Approximation and Projection, is a mathematical technique that takes very complex, high-dimensional data and simplifies it, making it easier to visualize and understand by squishing it down into fewer dimensions, often two or three.

Vision-Language, often abbreviated as VL, refers to the field of AI that focuses on building systems capable of understanding and processing both visual information (like images and videos) and textual information (like natural language). It's about teaching AI to see and to talk about what it sees.

In AI, especially in reinforcement learning, the "world" refers to the environment or setting where an AI agent exists, observes information, and performs actions, influencing its learning and behavior.

AI stands for Artificial Intelligence. It's about designing computer systems that can do tasks that usually need human smarts, like understanding language or making decisions.

CPU stands for Central Processing Unit. It's the primary component of a computer that performs most processing functions, often called the 'brain' of the computer.

The Entity-Relationship Model, or ER Model, is a way to visually represent the structure of a database, showing how different pieces of information, called entities, relate to each other.

In AI, a feature is an individual, measurable property or characteristic of a phenomenon being observed, used by a machine learning model to learn.

GPU stands for Graphics Processing Unit. It's a specialized electronic circuit designed to rapidly manipulate and alter memory to accelerate the creation of images in a frame buffer for output to a display device.

Machine learning is a field of artificial intelligence that lets computers learn from data without being explicitly programmed for every task.

Mean Squared Error, or MSE, is a common way to measure the average squared difference between predicted values and actual values in a statistical model.

The Minimax algorithm is a decision-making rule used in artificial intelligence, especially for two-player games, that helps an AI pick the best move by assuming the opponent will play optimally to minimize the AI's gains.

A neural network is a type of computer program inspired by the structure and function of the human brain, designed to recognize patterns and make decisions.

Open Source Software, or OSS, is software whose source code is made publicly available for anyone to use, modify, and distribute freely.

In AI, a parameter is a setting within a model that it learns and adjusts during training. These settings determine how the model processes information and makes predictions.

PCA stands for Principal Component Analysis. It's a fundamental statistical technique used to simplify complex datasets by reducing the number of variables.

Perplexity is a way to measure how well a language model predicts a sequence of words. A lower perplexity score means the model is better at predicting the next word, which indicates it "understands" the text more accurately.

Reinforcement Learning (RL) is a type of machine learning where an AI learns to make decisions by trying things out and getting rewards or penalties.

SVD, or Singular Value Decomposition, is a powerful mathematical technique for breaking down a matrix into three simpler matrices. It helps us understand and simplify complex data.

SSL, or Secure Sockets Layer, is a foundational technology that creates an encrypted link between a web server and a web browser, ensuring that all data passed between them remains private.

Tokens are small pieces of text or code that large language models (LLMs) use to understand, process, and generate language, like the building blocks of words.

TPS, or Transactions Per Second, is a performance metric that measures how many operations or requests a computer system, such as a database or an API, can process and complete within one second. It's a key indicator of a system's capacity and throughput.

AIME stands for Artificial Intelligence in Medical Epidemiology, a technology that uses AI to analyze and predict disease patterns.

AlphaFold is an AI system developed by Google's DeepMind that predicts the 3D structure of proteins from their amino acid sequence. This prediction is crucial for understanding how proteins work and for drug discovery.

AutoGen is an open-source framework developed by Microsoft that helps you build powerful AI applications by allowing multiple AI agents to chat and collaborate on a task.

A CodeAgent is an AI system designed to help developers write, debug, and improve code. It acts like an intelligent assistant, understanding code, finding errors, and suggesting solutions.

CRM stands for Customer Relationship Management. It's a system businesses use to manage interactions with their current and potential customers.

Cursor is an AI-native code editor designed to help developers write, debug, and understand code more efficiently by integrating large language models directly into the development workflow.

Descript is an all-in-one audio and video editing tool that uses AI to make editing as simple as editing a text document.

Devin is an AI software engineer developed by Cognition AI, designed to autonomously plan, execute, and debug entire software development projects from start to finish.

GGUF is a file format designed to store large language models in a way that's efficient for running them on your own computer, even without a super powerful graphics card.

GitHub is a popular web-based platform that helps developers store, manage, and collaborate on software projects using Git, a version control system.

HARPA AI is a browser extension that brings artificial intelligence capabilities directly into your web browser, allowing for automation, content generation, and smart web interaction.

HyperWrite is an AI-powered writing assistant that helps users generate, rewrite, and improve various types of text content.

An Integrated Development Environment, or IDE, is a software application that provides a complete set of tools to help programmers write, test, and debug their code efficiently.

JavaScript is a popular programming language that brings interactive features to websites, making them dynamic and engaging.

LangChain is an open-source framework that helps developers build applications powered by large language models (LLMs) by making it easier to connect LLMs with other data sources and tools.

LangGraph is a Python library that helps you build complex, multi-step AI applications by letting you define workflows as graphs, where each step is a node and arrows show the flow of information.

LangSmith is a development platform created by LangChain that helps developers build, debug, test, and monitor applications powered by large language models (LLMs). It provides tools to visualize the internal steps of an LLM application, making it easier to understand and improve its performance.

MindStudio is an integrated development environment (IDE) created by Huawei. It's designed to help developers build, optimize, and deploy artificial intelligence (AI) applications specifically for Huawei's Ascend series AI processors.

NAS stands for Network Attached Storage, which is a dedicated file storage device connected to your network that allows multiple users and devices to store and retrieve data from a central location.

Notion is a versatile workspace application that combines notes, tasks, wikis, and databases into one customizable platform, helping individuals and teams organize information and manage projects.

NVIDIA ACE is a set of tools for building AI-powered digital humans, and 'agents' refer to these intelligent virtual characters that can interact in lifelike ways.

NVIDIA DGX Systems are powerful, pre-built computing platforms designed specifically for artificial intelligence and machine learning workloads, from training large models to running complex simulations.

NVIDIA NeMo is a toolkit for developers to build, train, and deploy large language models and other generative AI models efficiently.

PyCharm is a popular software development environment specifically designed to help programmers write, debug, and run Python code more efficiently.

Ray is an open-source framework that helps you scale your Python code across multiple computers, making it easier to run complex AI and machine learning tasks faster.

RunDiffusion is a cloud platform that makes it easier to run popular AI image generation tools like Stable Diffusion and ComfyUI without needing powerful, expensive computer hardware of your own.

SVG stands for Scalable Vector Graphics. It is a file format for creating two-dimensional graphics that can be scaled up or down without losing quality.

TikTok is a popular social media app for creating and sharing short-form videos. Its success is largely driven by a sophisticated artificial intelligence algorithm that personalizes content recommendations for each user, creating its famous "For You Page."

TypeScript is a programming language that builds on JavaScript by adding optional static type definitions, helping developers catch errors early and write more reliable code.

A vector database is a special type of database designed to store, manage, and quickly search through information represented as 'vectors' or numerical lists. It's especially useful for finding similar items based on their meaning, not just keywords.

Vertex AI is Google Cloud's integrated platform that provides tools and services for developers and data scientists to build, deploy, and manage machine learning models throughout their entire lifecycle.

WebP is a modern image format developed by Google that offers superior compression for images on the web. It aims to make websites load faster while maintaining high image quality, supporting both lossy and lossless compression.

YAML (YAML Ain't Markup Language) is a human-friendly data format often used for configuration files and data exchange, making it easy to read and write structured information.

Zapier is an online automation tool that connects different web applications, allowing them to work together without requiring any coding. It helps you automate repetitive tasks across various services like email, spreadsheets, and social media.

Attention is a powerful mechanism in AI, especially in models like Transformers, that helps the model focus on the most important parts of its input.

CNN stands for Convolutional Neural Network, a type of neural network especially good at processing images and other grid-like data.

Graph Neural Networks (GNNs) are a type of neural network designed to work directly with data structured as graphs, like social networks or molecular structures.

GRU stands for Gated Recurrent Unit. It's a type of artificial neural network designed to process sequences of data, like words in a sentence or measurements over time.

Long Short-Term Memory, or LSTM, is a special kind of recurrent neural network designed to remember information for longer periods than standard networks.

An NPU, or Neural Processing Unit, is a specialized computer chip designed to speed up tasks related to artificial intelligence and machine learning.

A Recurrent Neural Network, or RNN, is a type of neural network that's really good at processing sequences of data, like text or time series, because it has an internal memory.

A TPU, or Tensor Processing Unit, is a specialized computer chip developed by Google specifically designed to accelerate artificial intelligence workloads, particularly the training and inference of neural networks.

A neural network architecture, or blueprint, especially good at understanding sequences of data like text or speech. It processes all parts of a sequence at once, rather than one piece at a time.

A Variational Autoencoder (VAE) is a type of neural network that learns to create new data, like images or text, by first compressing existing data into a structured 'code' and then using that code to generate similar, yet novel, outputs.

VLM stands for Vision-Language Model, which is a type of AI that can understand and connect information from both images and text.

Autonomy in AI means a system can make its own decisions and act independently to achieve a goal, without constant human intervention.

A technique that encourages AI models to show their step-by-step reasoning process, much like a human thinking aloud, to solve complex problems more accurately.

When an AI model generates information that sounds plausible but is factually incorrect, made-up, or inconsistent with its training data, it's called a "hallucination."

In AI, inference is the process where a trained model uses what it has learned to make predictions or generate outputs based on new, unseen data.

LIME is a technique that helps us understand why a complex AI model made a particular decision, especially for individual predictions.

MMLU stands for Massive Multitask Language Understanding. It's a popular benchmark used to evaluate how well large language models understand and answer questions across a wide range of academic and professional subjects.

Next Token refers to the fundamental step where a large language model predicts the most probable next piece of text, like a word or part of a word, in a sequence. This is how LLMs generate all of their output, building responses one piece at a time.

ReAct is a prompting technique that helps large language models (LLMs) reason and act more effectively by combining thought and action steps.

A reasoning model in AI is a system designed to simulate human-like thought processes, allowing AI to draw conclusions and make decisions based on information.

SHAP is a method that helps you understand why a machine learning model made a specific prediction by showing how much each input feature contributed to that outcome.

Temperature is a setting in AI models that controls how predictable or creative their responses are. A higher temperature means more randomness and variety, while a lower temperature makes responses more focused and consistent.

BERT stands for Bidirectional Encoder Representations from Transformers. It's a groundbreaking AI model developed by Google that revolutionized how computers understand human language by processing words in relation to all other words in a sentence.

Claude is a family of large language models developed by Anthropic, designed to be helpful, harmless, and honest. They are known for their strong performance in various tasks and their focus on ethical AI development.

Claude Sonnet is a powerful, mid-range artificial intelligence model from Anthropic, part of their Claude 3 family. It offers a great balance of performance and efficiency, making it suitable for a wide variety of business tasks.

CLIP, which stands for Contrastive Language-Image Pre-training, is an AI model developed by OpenAI that excels at understanding the relationship between images and text. It learns to connect visual information with its corresponding written descriptions across a huge variety of data.

DBRX is a powerful open source large language model developed by Databricks, known for its efficiency and strong performance across various tasks.

A diffusion model is a type of generative artificial intelligence that learns to create new data, such as images, audio, or video, by gradually refining random noise into a clear output.

Gemini is a family of powerful artificial intelligence models developed by Google that can understand and process different types of information, like text, images, and audio, all at once.

A large language model, or LLM, developed by Sber, a major Russian technology and financial company, designed to process and generate human-like text across various tasks and languages.

GPT stands for Generative Pre-trained Transformer. It is a type of large language model designed to understand and create human-like text, powering many AI applications today.

Llama is a family of powerful open-source large language models (LLMs) developed by Meta, designed to be used by researchers and developers.

LLM stands for Large Language Model. These are powerful computer programs, often called AI, designed to understand, generate, and respond to human language.

Nemotron is a family of large language models developed by NVIDIA, designed to power various AI applications from chatbots to content generation. They come in different sizes, optimized for diverse tasks and computing needs.

Qwen2 is a powerful family of large language models developed by Alibaba Cloud, known for its strong performance, multilingual capabilities, and open-source availability in various sizes.

A Small Language Model (SLM) is a type of language model with fewer parameters and less training data compared to its larger cousins, like GPT-4 or Claude Opus.

StarCoder2 is a large language model designed specifically for understanding and generating computer code across many programming languages. It's an open-source model that can help with tasks like writing new code, completing existing code, and even explaining code.

CutMix is a data augmentation technique used to train image classification models more effectively by mixing parts of different images together.

Fine-tuning is a technique where you take a pre-trained artificial intelligence model and adapt it for a more specific task or dataset by continuing to train it with new, specialized data.

HRL is a technique in AI where an intelligent agent learns to solve complex problems by breaking them down into smaller, more manageable sub-problems, much like how a human tackles a big project by setting intermediate goals.

Multi-Task Learning is a way to train a single AI model to perform several related tasks at once, helping it learn more efficiently and perform better on all of them.

PEFT, or Parameter-Efficient Fine-Tuning, is a collection of techniques that allow you to adapt large AI models to new tasks with much less computational power and storage than traditional fine-tuning.

RLHF is a training technique that uses human preferences to fine-tune large language models, making them more helpful, honest, and harmless.

SGD is a popular optimization algorithm used to train machine learning models, especially neural networks. It helps models learn by making small, iterative adjustments to their internal settings based on training data.

Training is the process where an AI model learns from data to identify patterns and make predictions or decisions.

Elon Musk is a well-known entrepreneur and CEO of several prominent technology companies, including Tesla, SpaceX, and X (formerly Twitter). He has a significant impact on areas like electric vehicles, space exploration, and artificial intelligence.

FunSearch is a computer program that uses large language models (LLMs) to help discover new mathematical functions and algorithms.

A GAN, or Generative Adversarial Network, is a type of artificial intelligence system made of two competing neural networks that learn to create realistic new data, like images or audio, that didn't exist before.

Prompt engineering is the art and science of crafting the best possible instructions or questions (prompts) to get the most helpful and accurate responses from an AI model.

GPQA is a very challenging benchmark designed to test the limits of advanced AI models in complex, multi-step reasoning across various scientific domains.

OmniDocBench is a benchmark designed to evaluate how well large language models (LLMs) can understand and process very long and complex documents, especially those found in real-world scenarios.

Model Context Protocol. An open standard from Anthropic for connecting AI assistants to external tools and data.

A prompt is the input you give to an AI, like a question or instruction, to tell it what you want it to do.

REST stands for REpresentational State Transfer. It's a widely used architectural style for building web services that allow different computer systems to communicate over the internet.

UI stands for User Interface, which is everything a person interacts with when using a digital product like an app or website. It includes visuals like buttons and text, and how you get around the application.

RAG stands for Retrieval Augmented Generation, a technique that helps large language models (LLMs) access and use up-to-date or specific information outside of their original training data.