Learn: the AI Roadmap and Encyclopedia as a feed
Everything on the AI Roadmap and in the AI Encyclopedia, in learning order.
Phase 1: Foundations
- What is AI? · The Big Picture (16 sources)
Phase 2: Python & Math
- Python Programming for AI (18 sources)
- Math for AI · Linear Algebra & Calculus (14 sources)
- Statistics & Probability for ML (16 sources)
- NumPy, Pandas & Data Visualization (14 sources)
Phase 3: Machine Learning
- Data Engineering & Data Pipelines for AI (17 sources)
- Intro to Machine Learning (17 sources)
- Model Evaluation & Experiment Tracking (15 sources)
- Scikit-Learn & Hands-On ML Projects (14 sources)
- 🧪 PRACTICE MILESTONE: Build Your First ML Project (12 sources)
Phase 4: Deep Learning
- Neural Networks & Deep Learning Fundamentals (19 sources)
- PyTorch · The Deep Learning Framework (17 sources)
- CNNs & Computer Vision (14 sources)
- RNNs, LSTMs & Sequence Models (12 sources)
- Transformers & Attention Mechanism (18 sources)
- 🏗️ PRACTICE MILESTONE: Deep Learning Projects (11 sources)
Phase 5: NLP & LLMs
- NLP & Text Processing Fundamentals (14 sources)
- Prompt Engineering & Using LLMs Effectively (18 sources)
- Large Language Models (LLMs) · GPT, BERT & Beyond (24 sources)
- Fine-Tuning & RAG · Customizing LLMs (19 sources)
- 🚀 PRACTICE MILESTONE: Build & Deploy an LLM App (14 sources)
Phase 6: Specialize & Build
- Generative AI · Images, Audio & Video (16 sources)
- AI Agents & Agentic AI (24 sources)
- MLOps & Deploying AI Models (15 sources)
- Reinforcement Learning Deep Dive (16 sources)
- Build Your AI Portfolio & Get Hired (16 sources)
- AI Safety, Ethics & Responsible AI (14 sources)
- Reading Research Papers & Staying Current (16 sources)
Phase 7: Depth Tracks
- Multimodal AI & Vision-Language Models (14 sources)
- Speech & Audio AI (11 sources)
- Robotics & Embodied AI (8 sources)
- Efficient AI · Quantization & Fast Inference (8 sources)
- Recommender Systems (6 sources)
- Time Series & Forecasting (11 sources)
- Graph Machine Learning (6 sources)
Core ML Concepts
- Artificial Intelligence (AI) (19 sources)
- Machine Learning (ML) (18 sources)
- Supervised Learning (12 sources)
- Unsupervised Learning (8 sources)
- Neural Network (24 sources)
- Deep Learning (20 sources)
- Features & Labels (6 sources)
- Overfitting & Underfitting (9 sources)
- Bias-Variance Tradeoff (6 sources)
- Classification vs Regression (7 sources)
- Transfer Learning (8 sources)
- Training vs Inference (5 sources)
- Regularization (8 sources)
- Few-Shot & Zero-Shot Learning (7 sources)
- Linear Regression (11 sources)
- Semi-Supervised Learning (6 sources)
- Time Series Forecasting (6 sources)
- Data · The Fuel of AI (3 sources)
- Decision Trees & Random Forests (12 sources)
- Support Vector Machines (SVM) (10 sources)
- Evaluation Metrics & Cross-Validation (10 sources)
- Self-Supervised Learning (7 sources)
- Continual Learning & Catastrophic Forgetting (6 sources)
- Meta-Learning (Learning to Learn) (5 sources)
- Causal Inference & ML (9 sources)
- Symbolic AI & Expert Systems (9 sources)
- Logistic Regression (12 sources)
- Data Leakage (6 sources)
- Exploratory Data Analysis (EDA) (9 sources)
- Data Augmentation (8 sources)
- k-Nearest Neighbors (kNN) (6 sources)
- Naive Bayes (7 sources)
- Data Science vs ML vs AI (11 sources)
- Narrow AI vs Strong AI (7 sources)
- Ensemble Learning (6 sources)
- Clustering (6 sources)
- Data Labeling (5 sources)
- History of AI (6 sources)
- Dropout (4 sources)
- Knowledge Representation & Reasoning (4 sources)
- Active Learning (5 sources)
- Data-Centric AI (13 sources)
- Data Quality for AI (15 sources)
- Data Hardness and Example Difficulty (12 sources)
- Hard Negative Mining (12 sources)
- Datasheets, Data Cards and Model Cards (13 sources)
- Survival Analysis (12 sources)
- Reward Models and Process Reward Models (14 sources)
- Curriculum Learning (10 sources)
Math & Optimization
- Gradient Descent (11 sources)
- Backpropagation (13 sources)
- Loss Function (6 sources)
- Activation Functions (8 sources)
- Embeddings (22 sources)
- Dimensionality Reduction (10 sources)
- Feature Engineering (13 sources)
- K-Means Clustering (10 sources)
- Anomaly Detection (8 sources)
- Boosting (XGBoost, LightGBM) (14 sources)
- Linear Algebra for ML (12 sources)
- Probability & Bayes' Theorem (13 sources)
- Information Theory: Entropy & KL Divergence (7 sources)
- Optimizers: SGD, Momentum & Adam (8 sources)
- Bayesian ML & Uncertainty (9 sources)
- Monte Carlo Methods (9 sources)
- Probabilistic Graphical Models & HMMs (8 sources)
- Evolutionary & Swarm Optimization (5 sources)
- Latent Space (4 sources)
- Statistics for ML (5 sources)
- Calculus for ML (4 sources)
- Search Algorithms (A*, Minimax) (7 sources)
- Fuzzy Logic (5 sources)
- Game Theory for AI (6 sources)
- Double Descent (14 sources)
- Grokking (12 sources)
- The Lottery Ticket Hypothesis (14 sources)
- Gaussian Processes (13 sources)
Architectures
- Transformer Architecture (19 sources)
- Attention Mechanism (13 sources)
- Self-Attention (11 sources)
- Convolutional Neural Network (CNN) (13 sources)
- Recurrent Neural Network (RNN) (10 sources)
- LSTM (Long Short-Term Memory) (11 sources)
- GPT (Generative Pre-trained Transformer) (12 sources)
- BERT (Bidirectional Encoder) (8 sources)
- State Space Models (Mamba) (7 sources)
- Vision Transformer (ViT) (9 sources)
- Multimodal Models (9 sources)
- Knowledge Distillation (7 sources)
- Model Quantization & Compression (8 sources)
- Neural Architecture Search (NAS) (7 sources)
- Neuro-Symbolic AI (5 sources)
- Mixture-of-Experts (MoE) (9 sources)
- Residual Connections & ResNet (5 sources)
- Normalization Layers (BatchNorm to RMSNorm) (6 sources)
- Autoencoders (7 sources)
- Graph Neural Networks (GNNs) (9 sources)
- Positional Encoding & RoPE (8 sources)
- Encoder-Decoder Models (5 sources)
- Multilayer Perceptron (MLP) (4 sources)
- Gated Recurrent Unit (GRU) (4 sources)
- Boltzmann Machines (5 sources)
- Spiking Networks & Neuromorphic Computing (5 sources)
- FlashAttention and IO-Aware Attention (14 sources)
- GQA, MQA, Multi-head Latent and Sliding Window Attention (14 sources)
- Diffusion Language Models (15 sources)
- Model Merging (14 sources)
- Pruning and Sparsity (13 sources)
NLP & Language
- Natural Language Processing (NLP) (12 sources)
- Tokenization (7 sources)
- Word2Vec & Word Embeddings (8 sources)
- Prompt Engineering (16 sources)
- Retrieval-Augmented Generation (RAG) (24 sources)
- Fine-Tuning LLMs (16 sources)
- Chain-of-Thought Reasoning (13 sources)
- Vector Databases (16 sources)
- LLM Scaling Laws (7 sources)
- RLHF (Reinforcement Learning from Human Feedback) (13 sources)
- Context Window & Long Context (9 sources)
- Text Generation & Sampling (7 sources)
- Sentiment Analysis (7 sources)
- Machine Translation (10 sources)
- Named Entity Recognition (NER) (6 sources)
- Reasoning Models & Test-Time Compute (17 sources)
- Context Engineering (10 sources)
- Speech Recognition (ASR) & Whisper (12 sources)
- DPO & Preference Optimization (5 sources)
- LoRA & Parameter-Efficient Fine-Tuning (6 sources)
- Open-Weights & Small Language Models (13 sources)
- Bag-of-Words & TF-IDF (6 sources)
- Text Summarization (5 sources)
- Topic Modeling (6 sources)
- Information Extraction (5 sources)
- Large Language Models (LLMs) (20 sources)
- Semantic Search (10 sources)
- Hybrid Search & Reranking (7 sources)
- Chunking (5 sources)
- Knowledge Graphs (7 sources)
- GraphRAG (6 sources)
- Structured Outputs (5 sources)
- Question Answering (6 sources)
- NLP Metrics (BLEU, ROUGE) (3 sources)
- Data Deduplication and Pruning (14 sources)
- Pretraining Data and Data Mixtures (15 sources)
- Token Density and Information Density (15 sources)
- Topic Density and Topical Coverage (13 sources)
- Prompt and Context Compression (14 sources)
- ColBERT and Late Interaction Retrieval (14 sources)
- Semantic Caching (13 sources)
- Text-to-SQL and Natural Language Querying (14 sources)
- Agentic RAG (14 sources)
Computer Vision
- Image Classification (9 sources)
- Object Detection (9 sources)
- Semantic Segmentation (9 sources)
- YOLO (You Only Look Once) (6 sources)
- Image Generation (11 sources)
- Neural Style Transfer (5 sources)
- Deepfakes & Face Synthesis (4 sources)
- Contrastive Learning & CLIP (7 sources)
- OCR & Document AI (9 sources)
- Face Detection & Recognition (6 sources)
- Computer Vision (10 sources)
Generative AI
- Synthetic Data (8 sources)
- Diffusion Models (18 sources)
- GANs (Generative Adversarial Networks) (11 sources)
- VAEs (Variational Autoencoders) (7 sources)
- Stable Diffusion & Text-to-Image (13 sources)
- Text-to-Video Generation (7 sources)
- Voice Cloning & TTS (6 sources)
- AI Music Generation (5 sources)
- AI Code Generation (6 sources)
- 3D Generation & NeRFs (6 sources)
- AI Image Editing & Inpainting (6 sources)
- Flow Matching & Modern Diffusion (6 sources)
- Generative AI (7 sources)
- Foundation Models (6 sources)
- Speech-to-Speech and Realtime Voice Models (14 sources)
- Any-to-Any (Omni) Models (13 sources)
Reinforcement Learning
- Reinforcement Learning (RL) (21 sources)
- Q-Learning & Deep Q-Networks (DQN) (9 sources)
- Policy Gradient & Actor-Critic (10 sources)
- PPO (Proximal Policy Optimization) (9 sources)
- Markov Decision Process (MDP) (8 sources)
- Reward Design & Reward Hacking (6 sources)
- Multi-Agent Reinforcement Learning (6 sources)
- World Models (7 sources)
- Sim-to-Real Transfer (4 sources)
- Inverse Reinforcement Learning (4 sources)
- GRPO (Group Relative Policy Optimization) (5 sources)
- Imitation Learning (5 sources)
- Offline Reinforcement Learning (14 sources)
- AlphaZero and MuZero (15 sources)
MLOps & Infrastructure
- GPU & TPU Computing (11 sources)
- Model Deployment & Serving (6 sources)
- MLOps (ML Operations) (10 sources)
- Experiment Tracking (4 sources)
- Data Pipelines & ETL (9 sources)
- Edge AI & On-Device ML (10 sources)
- Federated Learning (9 sources)
- Distributed Training (10 sources)
- Model Versioning & Registry (4 sources)
- ML Monitoring & Observability (6 sources)
- LLM Evaluation & Benchmarks (20 sources)
- Inference Optimization (10 sources)
- Hyperparameter Tuning (9 sources)
- The AI Stack (7 sources)
- Quantum Machine Learning (6 sources)
- AI Data Centers & the AI Factory (7 sources)
- Numerical Precision & Mixed-Precision Training (7 sources)
- ML Frameworks (PyTorch, TensorFlow, JAX) (10 sources)
- Inference Cost & Token Economics (12 sources)
- AI Platforms (11 sources)
- Digital Twins (4 sources)
- Prompt Caching (5 sources)
- LLMOps (6 sources)
- Model Routing & AI Gateways (4 sources)
- Feature Stores (5 sources)
- Model & Data Drift (5 sources)
- AI Energy & Environmental Impact (5 sources)
- AutoML (5 sources)
- A/B Testing for ML (5 sources)
- Ablation Studies (3 sources)
- Confidential Computing (4 sources)
- Batch vs Online Inference (3 sources)
- Gradient Checkpointing and Memory-Efficient Training (12 sources)
- Offline and Online Evaluation (12 sources)
AI Agents & Applications
- AI Agents (19 sources)
- Tool Use & Function Calling (13 sources)
- Autonomous Vehicles & Robotics (6 sources)
- Recommendation Systems (9 sources)
- AI-Driven Drug Discovery (4 sources)
- Explainable AI (XAI) (10 sources)
- AI-Powered Search & Ranking (7 sources)
- Conversational AI & Chatbots (9 sources)
- AI in Healthcare & Medical Imaging (7 sources)
- AI for Creative Industries (3 sources)
- Model Context Protocol (MCP) (13 sources)
- Computer Use & Browser Agents (5 sources)
- Agentic Coding (12 sources)
- Vision-Language-Action Models (6 sources)
- Multi-Agent Systems & Orchestration (11 sources)
- Agent Memory, Planning & Reflection (9 sources)
- Agent Harnesses & Scaffolding (7 sources)
- Agent-to-Agent Protocols (A2A) (10 sources)
- Enterprise AI (14 sources)
- Vibe Coding (14 sources)
- Agentic AI (9 sources)
- Physical AI & Robotics (5 sources)
- Agent Evaluation (5 sources)
- Agent Skills (5 sources)
- Automated Planning (4 sources)
- Cognitive Architectures (4 sources)
- Agentic Workflows & Patterns (4 sources)
- AI in Marketing & Search (5 sources)
- AI in Cybersecurity (4 sources)
- AI in Education (6 sources)
- AI in Finance (5 sources)
- AI for Science (13 sources)
- AI for Climate, Weather and Materials (15 sources)
- Levels of Agent Autonomy (13 sources)
- Agentic Commerce and Payments (14 sources)
- Voice Agents (15 sources)
Safety, Ethics & Governance
- AI Alignment (10 sources)
- AI Bias & Fairness (10 sources)
- AI Hallucinations (12 sources)
- Red Teaming AI Systems (5 sources)
- AI Regulation & Governance (10 sources)
- Prompt Injection & Jailbreaking (15 sources)
- Mechanistic Interpretability (13 sources)
- AI Watermarking & Content Detection (5 sources)
- Constitutional AI & RLAIF (4 sources)
- Differential Privacy & ML Privacy (5 sources)
- AGI (Artificial General Intelligence) (10 sources)
- AI Safety (14 sources)
- Superintelligence (ASI) (9 sources)
- Technological Singularity (10 sources)
- Responsible AI (14 sources)
- AI Guardrails (8 sources)
- Human in the Loop (5 sources)
- Adversarial Machine Learning (7 sources)
- Data Poisoning (7 sources)
- AI Governance (6 sources)
- AI Ethics (5 sources)
- AI Risk Management (4 sources)
- Deceptive Alignment & Scheming (4 sources)
- Frontier Safety Frameworks (4 sources)
- ARC-AGI and Reasoning Benchmarks (14 sources)
- Model Specs and Behavior Policies (14 sources)
- Scalable Oversight (13 sources)
- Existential Risk from AI (14 sources)
- AI Incidents and Incident Databases (12 sources)
- Children, Teens and AI (12 sources)
- AI, Elections and Information Integrity (12 sources)
- US AI Policy and Executive Orders (14 sources)
- Data Protection Law and AI (GDPR and beyond) (11 sources)
Business & Strategy
- AI Strategy (14 sources)
- Build, Buy or Partner (11 sources)
- Evaluating AI Vendors and Procurement (11 sources)
- AI Product Management (12 sources)
- Designing AI Products (Human-AI Interaction) (12 sources)
- AI Maturity Models (11 sources)
- AI Center of Excellence and Operating Model (11 sources)
- Data Strategy for AI (11 sources)
- AI Policy and Acceptable Use (13 sources)
- AI Adoption and Change Management (15 sources)
- Measuring AI Value and ROI (12 sources)
- AI and the Future of Work (13 sources)
- AI, Copyright and Training Data (12 sources)
- AI in Law and Legal Work (12 sources)
- AI in Sales and Customer Support (12 sources)
- AI in Retail, Supply Chain and Manufacturing (11 sources)