The Drum · Weekly AI news, freeSubscribe →
30 explainers
Plain-English breakdowns of the ideas, techniques, and terminology shaping AI. Each explainer covers what it is, why it matters, how it works, and what is commonly misunderstood.
AI bias explained
AI bias occurs when machine learning systems produce unfair or prejudiced outcomes due to flawed data or algorithms.
AI hallucination explained
AI hallucination occurs when AI systems generate outputs that are incorrect or inconsistent with reality.
AI safety explained
Explore the critical field of AI safety: understanding risks and solutions to ensure artificial intelligence benefits society.
AI watermarking explained
AI watermarking is a method to embed hidden information in AI-generated content, helping track and protect digital assets.
EU AI Act explained
The EU AI Act sets comprehensive rules for artificial intelligence use in Europe, shaping global standards.
How AI agents work
Dive into how AI agents operate, from decision-making to interaction, and their impact on industries.
How AI benchmarks work
Understand how AI benchmarks measure and compare the performance of different artificial intelligence models.
How AI chips work
Explore how AI chips process data at lightning speed to power complex machine learning models.
How AI coding assistants work
Discover how AI coding assistants streamline software development by offering real-time suggestions and error detection.
How AI inference pricing works
Understand how cloud providers charge for AI inference, a critical cost factor in deploying machine learning models.
How image generation works
Explore how AI turns text into vivid images, revolutionizing art and design.
How LLMs work
Explore how large language models (LLMs) process and generate human-like text, revolutionizing AI applications.
What are embeddings?
Embeddings transform textual and categorical data into numerical vectors, enabling machines to understand and process human language more effectively.
What is a foundation model?
Foundation models are large-scale AI systems trained on diverse data that can perform a wide range of tasks without extensive fine-tuning.
What is a vector database?
Explore how vector databases power advanced AI applications by efficiently storing and querying complex data.
What is agentic AI?
Agentic AI refers to artificial intelligence systems designed to act as autonomous agents, making decisions and interacting with complex environments.
What is AI alignment?
AI alignment focuses on ensuring that artificial intelligence systems do what we want them to do, safely and effectively.
What is constitutional AI?
Constitutional AI is a framework for ensuring that artificial intelligence systems operate within ethical and legal boundaries.
What is federated learning?
Federated learning is a privacy-preserving approach to machine learning that trains models on decentralized data without transferring it.
What is fine-tuning?
Fine-tuning is a technique in machine learning that adapts pre-trained models to perform specific tasks more accurately.
What is inference?
Inference is a critical process in AI that involves using trained models to make predictions or decisions based on new data.
What is LoRA?
Explore LoRA: a technique that enhances large language models with low-rank adaptation for efficient and targeted learning.
What is model distillation?
Model distillation is a technique that simplifies complex AI models to make them faster and more efficient without losing performance.
What is multimodal AI?
Multimodal AI combines data from multiple sources to enhance machine learning models, enabling more human-like understanding and interaction.
What is open-source AI?
Explore the world of open-source AI: what it is, how it works, and why it matters for innovation and accessibility.
What is prompt engineering?
Discover how prompt engineering shapes AI outputs and its significance in enhancing technology's usability.
What is RAG?
RAG is a retrieval-augmented generation model that combines information retrieval and language generation to improve AI's ability to answer questions accurately.
What is RLHF?
RLHF is a training method that teaches AI models to align better with human preferences and intentions.
What is synthetic data?
Explore how synthetic data is created and its crucial role in training AI models without compromising real-world privacy.
What is tool use / function calling?
Explore how AI systems can use external tools or functions to enhance their capabilities.
© 2026 Cedar & Bloom. All rights reserved.
