Basic Guide to Artificial Intelligence Terminology, Applications, and Examples

While tuning in to an AI podcast recently, I found the abundant use of abbreviations quite overwhelming. To help those who are still navigating the world of AI, I've compiled a user-friendly glossary covering some of the most frequently used terms in the field. Each term comes with a simple explanation, its potential applications, and an example to make the concepts more accessible and relatable. This glossary aims to demystify at least some of the AI jargon and provide a better understanding for newcomers.

GPT: (Generative Pre-trained Transformer) A powerful computer program developed by OpenAI that can understand and generate human-like text.

• Applications: Text generation, summarisation, translation, chatbots, and heaps more.
• Example: OpenAI's GPT-3 and 4 used in platforms like ChatGPT and various business applications. Some of this post was pulled together using ChatGPT-4!

AI: (Artificial Intelligence) is when computers can do tasks that usually require human thinking, like recognising patterns or solving problems.

• Applications: Image recognition, natural language understanding, data analysis, autonomous vehicles, and more.
• Example: Siri, Apple's some what helpfull personal assistant.

AGI: (Artificial General Intelligence) A type of AI that can learn and perform many different tasks, just like a human can. We haven't achieved this level of AI yet, and this is what a lot of people are mostly freaked out about.

• Applications: Once developed, AGI could be applied to virtually any field or task.
• Example: No real-world examples exist yet, as AGI remains a theoretical concept, but basically everything.

LLM: (Large Language Model) A computer program that can understand and work with human language, often used for tasks like writing or answering questions.

• Applications: Text summarisation, sentiment analysis, language translation, and more.
• Example: Google's BERT used in search engine algorithms and natural language understanding tasks.

ML: (Machine Learning) A way of teaching computers to learn from data and make decisions or predictions based on what they've learned.

• Applications: Fraud detection, personalised recommendations, predictive maintenance, and more.
• Example: Netflix's recommendation system for suggesting movies and TV shows.

NLP: (Natural Language Processing) The first time I heard this I thought it was Neuro Linguistic Programming. This is actually field of study that helps computers understand and communicate with people using human language.

• Applications: Sentiment analysis, voice assistants, text classification, and more.
• Example: Amazon Alexa, a voice-controlled virtual assistant.

ANN: (Artificial Neural Network)A type of computer program inspired by how the brain works, which can help solve complex problems.

• Applications: Image and speech recognition, text generation, and more.
• Example: Google DeepMind's AlphaGo, a program that defeated a world champion Go player.

LSTM: (Long Short-Term Memory) As confusing as it sounds, this is a kind of computer program designed to remember and process information that comes in a sequence, like sentences or time series data.

• Applications: Text prediction, speech recognition, time series forecasting, and more.
• Example: Gesture recognition systems for interpreting sign language.

RNN: (Recurrent Neural Network) A type of computer program that can process information that comes in a sequence, such as sentences or time-based data. Very similar to LSTM.

• Applications: Handwriting recognition, speech synthesis, and more.
• Example: Mobile phone keyboards with text prediction features.

CNN: (Convolutional Neural Network) Not the broadcaster - computer program that's especially good at understanding and processing images, like recognising objects or faces.

• Applications: Image classification, facial recognition, object detection, and more.
• Example: Facebook's DeepFace for facial recognition in photos.

BERT: (Bidirectional Encoder Representations from Transformers) Not a reference to Sesame Street. BERT is a powerful language program developed by Google that has helped improve how computers understand and generate human language. And no, there is no ERNIE (but there so should be!)

• Applications: Search engines, question-answering systems, sentiment analysis, and more.
• Example: Google Search using BERT for better understanding of search queries.

RL: (Reinforcement Learning) A way to teach computers to make decisions by interacting with their environment and learning from the feedback they get. Basically we give it treats for good treats or penalties.

• Applications: Game playing, robotics, autonomous vehicles, and more.
• Example: OpenAI's Dactyl, a robotic hand that can manipulate objects through reinforcement learning.

Hope you find this useful.