AI, GAI, ML, LLM, GANs, and GPTs: What are they ? The Inner Workings Made Simple
In the world of tech, we are surrounded by a sea of acronyms. For those not in the know, it can feel like trying to decode a secret message. Here’s a deeper, but still accessible, exploration of the technicalities of AI, GAI, ML, LLM, GANs, and GPTs.
AI, or “Artificial Intelligence,” is like the brain of a computer. It’s designed to mimic human thought processes, learning from experience, and adjusting to new inputs.
Just like how you learn to bake a cake by following a recipe, AI uses algorithms (the computer’s recipe) to learn from data.
A spam filtering AI learns to identify spam emails by analyzing thousands of emails that are labeled “spam” or “not spam.” Over time, the AI learns the patterns and characteristics that differentiate spam from non-spam.
GAI, or “Generative Artificial Intelligence,” is AI’s creative cousin. If AI is like following a recipe, GAI is like being a master chef, creating new recipes from scratch.
It can generate completely new content, like a song or an image, by learning patterns from existing data. It’s as if you trained a robot to understand music, and then it composed a completely new symphony!
A GAI trained on images of faces could generate a new image of a face that doesn’t belong to any real person, but still looks convincingly human. This is because the GAI has learned the fundamental features that make up human faces, such as the placement of eyes, nose, and mouth, and can recreate these features in new combinations.
ML, or “Machine Learning,” is the method AI uses to learn, kind of like how you learn by studying. It’s a type of AI that allows computers to learn from data without being explicitly programmed.
ML works by using algorithms to parse data, learn from it, and then make a determination or prediction about something. Machine learning involves feeding these algorithms (also known as models) with large amounts of data, which the model uses to improve its performance over time
For instance, a machine learning model could learn to identify cats in pictures by studying lots of cat pictures. It’s like learning to recognize your friend’s face in a crowd; after seeing it many times, you can pick it out even in a busy place.
LLM, or “Language Models,” are the linguists of the AI world. They study text data and learn to generate human-like text.
Just like how you learned language as a kid by listening to people talk, these models learn from reading huge amounts of text.
They generate human-like text by ‘filling in the blanks’ in a sequence of text. For instance, if you input the phrase “The weather today is,” the LLM might complete it as “The weather today is sunny and warm,” based on the patterns it has learned from the training data.
So if you feed it a lot of Shakespeare, it might start sounding like a 16th-century playwright!
GANs, or “Generative Adversarial Networks,” are two AI models that work together by competing against each other, like two artists trying to outdo each other. One creates (or ‘generates’) new data, and the other judges how good it is. This process improves the quality of the generated data over time. It’s like having an art contest between two painters, where one creates a painting and the other critiques it.
GANs can be used to create realistic images of people who don’t exist.
GPTs, or “Generative Pretrained Transformers,” are the storytellers of the AI world. They’re a type of LLM that can generate very convincing text. It’s like if you asked a skilled actor to pretend to be you; they’d study how you speak and what you talk about, and then they could make up a story that sounds just like something you’d say.
For example, GPT-4
So there you have it — AI, GAI, ML, LLM, GANs, and GPTs, decoded! These acronyms might sound intimidating, but when you break them down, they’re just different ways that computers are learning to act a bit more like us humans.
