Hardware-first basics: computers, chips, memory, networks, sensors, and models.
Hardware and software, from NIST: a machine that works with information.
Chips need electricity. ENERGY STAR measures power in Off, Sleep, and Idle. Heat is qualified.
RAM is temporary and lost if power is lost. Storage media can retain data; that is why sanitization exists.
RFC 791 datagrams go from source to destination. IP has no reliability, acks, or retransmissions.
Sensors turn a physical property into a voltage or current. Data is a limited record, not the whole event.
An AI model is a component that produces outputs from inputs. It does not replace the computer.
Core concepts everyone should know.
The current State of AI release: frontier models, reasoning agents, open weights, inference economics, safety, and governance.
A concise introduction to Artificial Intelligence — core ideas and where it shows up.
Search, rules, reasoning, planning, and optimization sit beside learning. An LLM is one kind of system, not the definition of AI.
A model does not learn reality. It learns statistical structure available through particular data, objectives, and feedback.
Words are not tokens. Tokens are not meaning. A context window is not memory. Model parameters are not the chat.
A number that looks like 80% sure is a score until someone shows it is calibrated. Fluent language is not confidence.
A demo proves possibility. An evaluation estimates performance. Neither automatically proves usefulness in your environment.
The deployed system is the human and the machine together. Automation, assistance, and human control are design choices, not a vibe.
Practical strategies for understanding and working with AI in everyday contexts.
Practical strategies for using AI tools effectively in your daily work and creative projects.
A claim is about a system, not a vibe. Tell a first-party fact from a rumor, a model from a product, and a benchmark from a capability.
Exploring the nature of intelligence, both human and artificial.
From symbolic systems to deep learning and beyond — a brief history.
Machine learning and deep learning fundamentals.
Key concepts: datasets, features, training/validation, bias-variance, and generalization.
Understanding how neural networks learn and make decisions.
How models learn: gradients, loss functions, and optimizers.
Scaling neural networks: architectures, representation learning, and training dynamics.
Neural network architectures powering modern AI.
Convolutions, receptive fields, and modern vision architectures.
Sequential modeling: recurrence, gating, and sequence learning.
Attention mechanisms, self-attention, scaling laws, and the transformer revolution.
How models like GPT and Claude understand and generate text.
Creating content with AI: images, text, and beyond.
How diffusion-based generative models work and how to use them.
Principles and patterns for effective prompting across tasks.
Combining search and generation for grounded outputs.
Adapting models and aligning behavior with human feedback.
Mathematical foundations and advanced theory.
The mathematical foundations powering modern AI - linear algebra, calculus, and probability explained at three levels.
A hardware-first journey from transistors and programs to machine-based AI systems. Machine learning is one family of methods — not the definition of AI.
Agents, environments, policies, value functions, and deep RL.
Safety, ethics, and where AI is headed.
Risks, safety frameworks, and alignment strategies.
Responsible AI: fairness, accountability, transparency, and governance.
Agent architectures, planning, tool use, and evaluation.
Scenarios, economics, governance, and societal impact.
Explore the interactive knowledge graph to visualize relationships between AI concepts. Perfect for understanding the big picture.
Explore Knowledge Graph