AI Glossary
A comprehensive guide to essential AI terminology, designed to make artificial intelligence concepts accessible to everyone.
Core AI Terms
1. Artificial Intelligence (AI)
Definition: Computer systems designed to perform tasks that typically require human intelligence, such as understanding language, recognizing patterns, making decisions, and solving problems.
In Practice: AI powers voice assistants like Siri and Alexa, recommendation systems on Netflix and Amazon, and autonomous vehicles. AI systems learn from data and experience rather than following only explicit programmed instructions.
Key Point: AI is an umbrella term covering many technologies, from simple rule-based systems to advanced machine learning models.
2. Machine Learning (ML)
Definition: A subset of AI where computers learn patterns from data without being explicitly programmed for every scenario. The system improves its performance as it processes more data.
In Practice: Email spam filters learn to identify spam by analyzing thousands of examples. Fraud detection systems recognize suspicious transactions by learning from historical patterns.
Key Point: Instead of writing rules for every possible scenario, ML systems discover patterns automatically from examples.
3. Large Language Model (LLM)
Definition: Advanced AI models trained on vast amounts of text data to understand and generate human-like language. These models can write, summarize, translate, answer questions, and perform various language tasks.
In Practice: ChatGPT, Claude, and Gemini are LLMs that can have conversations, write code, analyze documents, and assist with creative tasks. They process language by understanding context and relationships between words.
Key Point: LLMs don't "know" facts like a database; they generate responses based on patterns learned from training data.
4. Natural Language Processing (NLP)
Definition: The branch of AI focused on enabling computers to understand, interpret, and generate human language in a meaningful way.
In Practice: NLP powers chatbots, language translation services (like Google Translate), sentiment analysis of customer reviews, and voice-to-text systems.
Key Point: NLP bridges the gap between human communication and computer understanding, making human-computer interaction more natural.
5. Deep Learning
Definition: A specialized type of machine learning that uses artificial neural networks with multiple layers to learn increasingly complex patterns from data.
In Practice: Deep learning enables facial recognition in photos, voice recognition in smart speakers, and self-driving car vision systems. It's particularly effective for image, speech, and language tasks.
Key Point: "Deep" refers to the multiple layers in the neural network, each learning different levels of abstraction from the data.
6. Neural Network
Definition: A computing system inspired by biological brain neurons, consisting of interconnected nodes (artificial neurons) organized in layers that process and transform information.
In Practice: Neural networks power image recognition, speech synthesis, and pattern detection. Each layer of the network learns to recognize different features, from simple edges to complex objects.
Key Point: While inspired by the brain, artificial neural networks are mathematical models that work quite differently from biological neurons.
7. Training Data
Definition: The collection of examples, information, and content used to teach an AI system how to perform its task. The quality and diversity of training data directly impacts the AI's performance.
In Practice: An image recognition system is trained on thousands of labeled photos. A language model is trained on books, articles, and web content. Medical AI is trained on patient records and diagnostic images.
Key Point: Biased or limited training data leads to biased or limited AI capabilities. "Garbage in, garbage out" applies to AI training.
8. Prompt
Definition: The input text or instruction given to an AI system (especially LLMs) to generate a response or perform a task. Effective prompting is crucial for getting useful AI outputs.
In Practice: Asking ChatGPT "Write a professional email declining a meeting" is a prompt. More detailed prompts ("Write a polite, 3-sentence email declining next week's meeting due to schedule conflicts") typically yield better results.
Key Point: Prompt engineering—crafting effective prompts—is an emerging skill for working with AI systems effectively.
9. Fine-tuning
Definition: The process of taking a pre-trained AI model and further training it on specific, specialized data to adapt it for particular tasks or domains.
In Practice: A general language model might be fine-tuned on medical literature to better understand and generate medical content, or on legal documents to assist with legal research.
Key Point: Fine-tuning is more efficient than training from scratch and allows customization of AI models for specific use cases.
10. Hallucination
Definition: When an AI system generates information that sounds plausible but is actually incorrect, fabricated, or inconsistent with reality. A critical limitation to be aware of when using AI.
In Practice: An LLM might confidently cite non-existent research papers, create fake statistics, or describe events that never happened—all while presenting the information convincingly.
Key Point: Always verify important information from AI systems against reliable sources. AI confidence doesn't equal accuracy.
Using This Glossary
This glossary is designed to be:
- Accessible: Plain language explanations without assuming technical background
- Practical: Real-world examples showing how concepts apply
- Honest: Including both capabilities and limitations
- Evolving: Updated as AI technology and terminology develop
For more detailed information on any term, or to suggest additions, please refer to the project documentation.