01
AI Foundations
Understand what AI is, how systems learn, where they fail, and how to judge claims without hype.
Free AI library
Learn AI.Free forever.
1019 plain-English guides, structured learning paths, and an open library — built by an independent 501(c)(3) nonprofit so anyone can understand modern AI.
1019Free guides
9Topic tracks
~2 minPer guide
~34hReading time
Start here
Five outcome-based courses
Each course includes explicit outcomes, mapped competencies, practice activities, and an applied capstone.
02
Responsible AI User
Use AI productively while protecting privacy, checking outputs, and preserving human accountability.
03
AI at Work
Evaluate workplace use cases, run safe pilots, measure value, and communicate changes responsibly.
04
AI Policy & Society
Analyze AI systems through rights, equity, governance, safety, and public-interest outcomes.
05
Building with AI Systems
Understand language models, retrieval, agents, evaluation, cost, and deployment safeguards through practical system design.
Topic tracks
Browse by track
Jump into the area you care about. Every track has multiple plain-English guides.
FundamentalsCore concepts: how AI works at a high level.84 guides Language AILLMs, NLP, ChatGPT, prompting, and text understanding.155 guides Visual AIImage generation, computer vision, video, and design.133 guides Audio AIVoice, music, podcasting, and audio-first AI.117 guides CompaniesProfiles of OpenAI, Anthropic, Google, Meta, and more.121 guides ApplicationsHow AI shows up in coding, search, agents, and daily life.111 guides IndustriesAI in healthcare, finance, education, energy, and more.91 guides TechnicalChips, fine-tuning, RAG, vectors, and AI infrastructure.177 guides SocietyBias, privacy, copyright, safety, and the AI future.30 guides
Full library
All guides
1019 of 1019 guides shown. Filter by track or search above.
Technical
Consistency Regularization in Semi-Supervised Learning
Consistency regularization teaches a model to give the same answer when an unlabeled input is perturbed in small, label-preserving ways.
TechnicalHard Parameter Sharing in Multi-Task Networks
Hard parameter sharing is the classic multi-task learning design where several tasks share the same hidden layers and only split into separate output 'heads'…
TechnicalGating and Routing in Conditional Computation
Gating and routing let a neural network activate only the parts it needs for each input instead of running the whole model every time.
TechnicalGumbel-Softmax and Reparameterization
Gumbel-Softmax is a trick that lets neural networks 'sample' from discrete categories while still being trainable by gradient descent.
TechnicalStraight-Through Estimator
The Straight-Through Estimator (STE) is a simple trick for training networks that contain hard, non-differentiable steps like rounding or thresholding.
FundamentalsGroup Normalization
Group Normalization is a technique that stabilizes neural network training by normalizing features within small groups of channels, independently for each…
FundamentalsGated Recurrent Units
A Gated Recurrent Unit (GRU) is a streamlined type of recurrent neural network cell that uses two gates to decide what information to keep and what to forget…
TechnicalBidirectional Recurrent Networks
A bidirectional recurrent network reads a sequence both forward and backward, so each position's representation draws on context from the past and the future.
TechnicalAttention Rollout and Head Pruning
Attention rollout is a method for tracing how information flows through a Transformer's stacked attention layers to explain which input tokens influence…
Language AIContrastive Decoding
Contrastive decoding generates higher-quality text by subtracting the tendencies of a small, weak language model from those of a large, strong one.
Language AIGuided Beam Search with Constraints
Constrained beam search forces a language model's output to satisfy hard requirements, like including specific words or matching a grammar, while still…
Language AIMinimum Bayes Risk Decoding
Minimum Bayes Risk (MBR) decoding picks the output that is most similar to many other likely outputs, rather than the single highest-probability one.
Finished reading? Prove it.
Check what you learned with topic quizzes, then explore our structured courses and current certification requirements. Core guides remain free to read.