| 1. AI (Artificial Intelligence) |
The broad concept of machines acting in a way that simulates human intelligence. |
| 2. ML (Machine Learning) |
A subset of AI where computers learn from data patterns instead of strict rules. |
| 3. DL (Deep Learning) |
A specialized type of ML that uses "neural networks" to process complex data. |
| 4. LLM (Large Language Model) |
An AI (like GPT-4) trained on massive text data to understand and generate language. |
| 5. GenAI (Generative AI) |
AI that creates *new* content—text, images, video, or music—rather than just analyzing it. |
| 6. AGI (Artificial General Intelligence) |
A theoretical AI that can perform *any* intellectual task a human can do. |
| 7. NLP (Natural Language Processing) |
The tech that helps computers understand, interpret, and respond to human language. |
| 8. Prompt |
The specific instruction or question you give an AI to get a response. |
| 9. Prompt Engineering |
The art/skill of refining prompts to get the best possible output from an AI. |
| 10. Hallucination |
When an AI confidently provides information that is factually incorrect or nonsensical. |
| 11. RAG (Retrieval-Augmented Generation) |
A method where AI looks up specific, real-world data before answering to reduce errors. |
| 12. RLHF (Reinforcement Learning from Human Feedback) |
Training AI by having humans rank its answers, helping it align with human values. |
| 13. Neural Network |
A computer system designed to mimic the structure and function of the human brain. |
| 14. Algorithm |
A set of step-by-step instructions or rules that the AI follows to solve a problem. |
| 15. Training Data |
The massive "textbook" of information used to teach an AI model. |
| 16. Token |
The basic unit of text (like a word or part of a word) that an LLM processes. |
| 17. Parameters |
The internal "settings" or "knobs" an AI adjusts during training to learn patterns. |
| 18. Inference |
The process of the AI actually generating an answer based on what it learned. |
| 19. Context Window |
The amount of information (memory) the AI can "remember" during a single conversation. |
| 20. Temperature |
A setting that controls how "creative" (high) or "predictable" (low) the AI's output is. |
| 21. Bias |
Systematic errors where AI shows favoritism or prejudice based on its training data. |
| 22. XAI (Explainable AI) |
AI designed so humans can understand *why* it made a specific decision. |
| 23. CV (Computer Vision) |
AI's ability to "see" and interpret information from images or videos. |
| 24. AI Agent |
An AI system that can take independent actions (like booking a flight) to reach a goal. |
| 25. Edge AI |
AI that runs locally on a device (like your phone) instead of in the cloud. |
| 26. Guardrails |
Restrictions put on AI to keep it safe, ethical, and helpful. |
| 27. Deepfake |
AI-generated images, videos, or audio that look and sound like a real person. |
| 28. Fine-Tuning |
Taking a pre-trained AI and giving it extra training on a specific topic. |
| 29. GAN (Generative Adversarial Network) |
Two AI systems competing against each other to create highly realistic data/images. |
| 30. Zero-Shot Learning |
When an AI can perform a task it wasn't specifically trained for. |
