Skip to main content
What Is All This AI Stuff? Welcome! At Gloo AI, we believe everyone deserves to understand how modern AI works, without needing a PhD or building your own app. This lesson walks you through the core ideas behind everything from ChatGPT to image generators and smart search engines.

Artificial Intelligence (AI)

What it means: AI is the broad field of teaching machines to do things that typically require human intelligence—like understanding language, recognizing images, making decisions, or learning from data. Analogy: Think of AI as the umbrella term. It’s like saying “sports”—under that, you have soccer, tennis, basketball. In AI, those subfields include things like machine learning and computer vision. How it shows up in Gloo: Gloo AI gives organizations access to AI models through one API, and helps them use AI with their own content. The Data Engine makes your content searchable and lets AI answer from it, and Studio is where you try models and manage your account.

Machine Learning (ML)

What it means: Machine Learning is a branch of AI where computers learn patterns from data instead of being explicitly programmed: we feed the machine examples, and it figures out how to make predictions or decisions. Example: You don’t teach a machine every rule for recognizing cats in photos. Show it 10,000 pictures labeled “cat” and “not cat.” It learns the patterns that typically define a cat.

Deep Learning

What it means: Deep learning is a type of machine learning that uses neural networks with many layers to learn complex patterns. It works especially well for language, images, and audio. Analogy: If machine learning is like learning to recognize apples by color and shape, deep learning works out its own clues, like stem angle, skin texture, or shine, without being told which features to look for.

Neural Network

What it means: A neural network is a system loosely inspired by how the brain processes information. It’s made of connected units, often called “neurons,” that each process a signal and pass it along. Visualize it: Picture rows of dots (the neurons) connected by lines, each passing its signal to the next. The final output is the decision or prediction the system makes.

Transformer

What it means: A transformer is a type of neural network design that changed how machines handle sequences, like the words in a sentence. Why it matters: Before transformers, AI struggled with long text and with meaning that spans several sentences. Transformers use a technique called attention to weigh the most relevant parts of the input, not just the most recent word. Famous example: ChatGPT and nearly every powerful text model today is built on transformer architecture.

Diffusion

What it means: Diffusion is a technique used in image and video generation. The model starts with pure noise and gradually “denoises” it into a coherent image based on a prompt. Analogy: Imagine taking a photo and gradually adding static until it’s just noise. Diffusion models learn to reverse that process, ending with an image of, say, “a dog riding a skateboard in space.”

Model

What it means: A model is what you get after training an AI system. It’s a large set of learned numbers (called parameters) that turns an input, like a sentence or image, into an output, like a summary or a label. How to think about it: If data is the fuel, and algorithms are the engine, then the model is the car you build from both. It’s what you deploy when your AI is ready for real-world work. How it shows up in Gloo: Gloo AI gives you one API to models from OpenAI, Anthropic, Google, and open-weight providers. You can call a model directly, or through Gloo’s guarded endpoints, which add guardrails, output moderation, and values-aligned (tradition) responses.

Large Language Model (LLM)

What it means: A large language model (LLM) is an AI model trained on huge amounts of text so it can read and write natural language. How it works: It doesn’t “know” facts the way people do. It has learned patterns in language, such as which words are likely to come next, and uses them to write responses that can be surprisingly smart. Example: GPT, Claude, Gemini, and Llama are all families of LLMs. How it shows up in Gloo: LLMs power most of what you can do on Gloo AI: chat and responses, tool use, reasoning, and answers grounded in your own content. Go deeper: How AI Thinks shows how an LLM predicts text one token at a time.

Multimodal AI

What it means: Multimodal AI can work with more than one kind of input or output, such as text, images, audio, or video, in the same system. Use case: You can upload a picture and ask, “What’s happening in this photo?” and the AI can understand and respond. Or you might give it a chart and ask it to write a summary. How it shows up in Gloo: On the Responses endpoints, the same request can include an image for the model to read, or ask an image-capable model to generate one. The Data Engine also accepts content in many formats, such as files, websites, YouTube channels, and feeds, and makes it searchable. Go deeper: Images In and Out shows how to send images to a model and generate them.

Artificial General Intelligence (AGI)

What it means: Artificial general intelligence (AGI) is a hypothetical AI that could do any intellectual task a person can. Whether and when it will arrive is widely debated. Why it matters: Today’s AI is narrow: it does specific things well, like answering questions or generating images. AGI would be flexible, able to pick up new tasks on the fly like a person.
Try it on Gloo: browse the models you can call on Supported Models, then send your first request with the Developer Quickstart.

What’s next

When you’re ready to build, Supported Models lists the models you can call, and Choosing, Pinning and Updating Models helps you pick one.

How AI Models Learn

How do AI models learn, improve, and adapt after they’re built?