AI is a new and powerful way to program computers. This foundational article introduces the concept that neural networks are a different kind of programming paradigm.

Approachable explanation of how ChatGPT and GPT models work, how to use them, and where research and development is headed.

Long but highly readable explanation of modern AI from first principles. Wolfram breaks down the mechanics of language models in an accessible way.

Direct introduction to large language models with intuitive explanations of how transformer architecture works.

A layperson's guide to text-to-image models. Explains the mechanics of Stable Diffusion in an approachable way.

Four-part series covering deep learning fundamentals including neural networks, training, and optimization.

Comprehensive, free course teaching practical deep learning with code examples. Designed for developers who want to build real applications.

Introduction to embeddings and how words are represented as vectors in large language models.

In-depth explanation of backpropagation fundamentals. Essential for understanding how neural networks learn.

Introduction to machine learning course covering supervised and unsupervised learning, best practices, and real-world applications.

Natural language processing course covering modern deep learning approaches to understanding and generating text.

Technical overview of transformer architecture with visual explanations. Essential reading for understanding modern LLMs.

Source code level understanding of transformers with PyTorch implementation. Line-by-line breakdown of the architecture.

Video walkthrough building GPT from scratch. Learn by implementing a character-level language model.

Introduction to latent diffusion models with visual explanations of how Stable Diffusion generates images.

Explanation of RLHF and how it makes large language models more predictable and aligned with human preferences.

Deep dive on RLHF state, progress, and limitations from one of the technique's pioneers.

Seminar series on transformers covering recent advances and applications across different domains.

Course covering both technical and non-technical aspects of large language models including capabilities, limitations, and societal impact.

Early explanation of modern LLM application stack. Tutorial for building a practical chatbot application.

Key challenges in building production LLM applications including evaluation, monitoring, and deployment.

The most comprehensive guide to prompt engineering with model-specific examples and best practices.

Conversational treatment of prompt engineering with practical examples from production use.

Security vulnerability description in LLM applications. Essential reading for building secure AI systems.

Definitive collection of guides and code examples for building with OpenAI models.

Vector search paradigm instruction covering embeddings, similarity search, and vector databases.

LLM orchestration layer and full stack reference for building complex AI applications.

Practical course for building LLM applications covering the full stack from prompts to production.

Guide to using open-source LLMs in the transformers library with practical examples.

Value accrual assessment across AI layers from infrastructure to applications.

GPU acquisition and computing resource strategy for AI companies.

Consumer market opportunities across sectors from content creation to personal productivity.

Overview paper that shaped the "foundation models" terminology and framework.

Annual roundup of breakthroughs, industry developments, regulation, and economics in AI.

The original transformer paper that revolutionized AI. Introduced the attention mechanism and transformer architecture.

First publicly available large language model with many active variants still in use today.

First GPT architecture paper introducing the generative pre-training approach.

GPT-3 paper describing decoder-only architecture and demonstrating few-shot learning capabilities.

InstructGPT paper introducing human-in-the-loop training for instruction following.

Latest OpenAI model technical specifications and capabilities evaluation.

Competitive open-source model demonstrating that smaller models can achieve strong performance.