Artificial Intelligence (AI) Glossary of Terms

All technology comes with its own array of mind-boggling terms and phrases. The AI world is no different. Here are the most common ones you may hear as a non-techie beginner to AI, explained to us like we were 7 years old :

AI Ethics

This is about making sure robots are nice and do good things. It's like teaching a robot to be a good friend and not do anything mean or bad.

AI Safety

This is about making sure robots don't cause any harm. It's like teaching a robot to be careful and not hurt anyone or break anything.

Algorithm

This is a set of instructions that the robot follows to do a task or learn something. It's like a recipe that tells the robot what to do.

AutoML - Automated Machine Learning

This is when a robot helps make other robots smarter. It's like if a robot could teach other robots how to do things all by itself.

AIaaS - AI as a Service

This is when people can use a robot's abilities through the internet. It's like being able to ask a really smart robot for help anytime you need it, just by using your computer or phone.

AGI - Artificial General Intelligence

This is when robots can do lots of different tasks just like humans. It's like having a robot friend that is as smart as a grown-up and can do all sorts of things.

ANI - Artificial Narrow Intelligence

This is when robots are good at doing just one thing. It's like having a robot that is really good at playing chess but can't do anything else.

Bias

This is when a robot might favor one thing over another without a good reason. It's like when you like one ice cream flavor more than another, but the robot shouldn't do that. We need to teach robots to be fair.

ChatGPT - Chat Generative Pre-trained Transformer

This is a special kind of talking robot that is really good at understanding and using human language. It's like a robot that you can talk to, and it will help you with your questions and problems.

Data

This is the information that the robot uses to learn. It's like the robot's food. The more data you give it, the better it can learn.

Data Privacy

This is about keeping people's information safe. It's like when you keep your secrets in a special place so no one else can find them. We need to make sure robots don't share people's secrets without permission.

Deep Fakes

These are fake pictures, videos, or sounds that are created by robots using special technology. Deep fakes can make it look like someone said or did something they didn't really do. It's like when you put on a mask and pretend to be someone else, but the robot does it using computers, and it can be very convincing. Deep fakes can be fun, but they can also be used for bad things, like spreading lies or tricking people. That's why we need to be careful and learn how to tell the difference between what's real and what's fake.

Deep Learning

This is a special kind of machine learning where the robot learns by using something called "neural networks" that are inspired by how our brains work.

Emergent Ability

This is when a robot learns to do something new or unexpected without being directly taught. It's like when you discover a new way to play with your toys just by trying different things. Emergent abilities can be exciting because they show us that robots can be creative and come up with their own ideas, but they can also be surprising and make us think about how to keep robots safe and under control.

Explainability

This is about making sure we can understand how a robot makes its decisions. Sometimes, robots use really complicated ways to learn and think, and it can be hard for humans to figure out why they did something. Explainability is important because it helps us trust the robot and make sure it's doing the right thing. It's like when your friend explains why they made a certain choice, so you can understand and trust them better.

Generative Adversarial Networks (GANs)

This is a special way of teaching robots, where two robots work together to learn. One robot creates things, like pictures or stories, while the other robot tries to figure out if they are real or made up. It's like a game between the robots, where they keep trying to trick each other and get better at their jobs. This helps the robots learn faster and create really cool things.

GLLMM aka Gollem-Class AI - Generative Large Language Multi-Modal Model

This is a really fancy robot brain that can understand and use many different types of "languages" like pictures, sounds, and even things like DNA. It's like having a super smart robot that can do lots of things at once.

IoT - Internet of Things

This is when everyday objects are connected to the internet and can talk to each other or to robots. It's like if your toy car could tell your video game how fast it went today.

LLM - Large Language Model

This is a big robot brain that can understand and use lots of words. It's like a really smart talking robot that can help you with many things.

Machine Learning

It's like teaching a robot how to do things by showing it lots and lots of examples. The robot learns from those examples and gets better over time.

Natural Language Processing (NLP)

This is when the robot learns to understand and use human language. It's like when you learn to read and write.

Neural Networks

They are like the robot's brain. They have lots of little parts called "neurons" that work together to help the robot learn and make decisions.

Prompt / Prompt Generation

A prompt is like a question or a task you give to the robot. The robot then follows your instructions to help you. A Prompt Engineer is someone who is really good at giving the robot these questions or tasks to get the best answers. It's like being a really good coach for the robot.

Reinforcement Learning

This is when the robot learns by trying different things and getting rewards or penalties. It's like when you learn to ride a bike and you get better because you want to avoid falling down.

Robotics

This is when the robot learns to move and interact with the world. It's like when you learn to ride a bike or play with toys.

Superintelligence

This is when a robot becomes much smarter than any human. It's like if a robot could solve problems and think of ideas that even the smartest grown-ups couldn't imagine. We need to be careful with superintelligence, so the robot doesn't do anything that could be bad for us.

Supervised Learning

This is when the robot learns by being shown the right answers. It's like when a teacher helps you learn by showing you the correct way to do things.

Training

This is when you teach the robot by showing it lots of examples. It's like when you practice soccer to get better at it.

Transhumanism

This is the idea of using technology, like robots or special gadgets, to make humans better and stronger. It's like if you could wear special shoes that help you run really fast or have a robot part that helps you remember everything. Transhumanism is about making people's lives better with the help of technology. However, it could also be used for bad things if people use these new powers in mean or harmful ways, like hurting others or cheating to get what they want. We need to be careful and make sure technology is used for good purposes and not for causing harm.

Transfer Learning

This is when a robot uses what it learned from one task to help with a different task. It's like when you learn to ride a bike, and it helps you learn to ride a scooter faster.

TTE - Time to Edit

This is how long it takes to fix something. It's like when you make a mistake in your writing, and you need to erase it and write it correctly. In the AI world, it's about how fast a human or a robot can fix something.

Unsupervised Learning

This is when the robot learns by figuring things out on its own. It's like when you explore and learn things by yourself without a teacher telling you what to do.