Foundations — Free AI Learning Track

Last updated: 2026-07-11

What these systems are and how they learn.

The Foundations track explains what generative AI systems are and how they learn. It is the shared vocabulary layer for everything else on this site: if you can describe tokens, parameters, context windows, and reward signals precisely, you can govern the systems built on them.

Four modules build that vocabulary in sequence. Generative AI 101 covers the two core architectures, autoregressive and diffusion, and the full system around the model, from data pipelines to serving metrics. AI Concepts 101 frames the shift to Software 3.0, where natural language becomes the programming interface and evaluations replace traditional tests. The LLM Concepts module works through tokens, embeddings, the three training phases, prompting techniques, and the five core failure modes. Reinforcement Learning 101 explains the agent-environment loop and the human feedback process that aligns modern models.

Read the modules in order if you are new to the field. Each one is a short, plain-language briefing aimed at practitioners and decision-makers, and each ends with the questions a leader should ask and the failures to watch for.

Modules in this track

Generative AI 101 (4 min)

Every design decision about a GenAI product starts with what kind of model it needs and what the model actually does under the hood.

Generative AI products are built on one of two foundational architectures. Autoregressive models such as GPT, Claude, Gemini, and Llama generate text one token at a time, which is why latency and cost scale with response length. Diffusion models such as Midjourney, Stable Diffusion, and Sora start from random noise and iteratively denoise it into a coherent image or video. Modern systems combine these with multimodal capabilities by mapping text and images into a shared vector space. The model is only one piece; data pipelines, training methods, evaluation, safety, and serving infrastructure complete the system.

AI Concepts 101 (4 min)

Software engineering has split into three eras, and leaders need vocabulary to steer teams through the current one.

Software 1.0 is hand-written logic. Software 2.0 shifted logic from code to curated datasets and neural networks. Software 3.0 turns natural language into the programming interface, making agents and vibe coding possible. In this new paradigm, a large language model becomes the central processing unit of an application, with context as its working memory, embeddings and vector databases as its persistent store, agent loops as its control flow, and tool calling as its hands. Evaluations replace traditional tests for non-deterministic behavior.

LLM Concepts, A Deep Dive (4 min)

The vocabulary for discussing language models is the vocabulary for governing them.

A large language model is an advanced autocomplete system. It predicts the next token in a sequence, one at a time, until it generates a stop signal. Its internal machinery has three layers worth knowing: tokens, embeddings, and parameters. Its training has three phases: pre-training on broad data, fine-tuning on a narrow dataset, and alignment through reinforcement learning from human feedback. Its interaction surface has two parts: the system prompt that shapes behavior and the user prompt that delivers the task. Its deployment has tradeoffs across proprietary, open-weight, and open-source variants, across small and large sizes, and across single-modality and multimodal designs.

Reinforcement Learning 101 (4 min)

Reinforcement learning now sits behind the models and agents senior leaders are asked to approve.

Reinforcement learning is a branch of machine learning in which an agent interacts with an environment, takes actions, receives rewards or penalties, and gradually learns a policy that maximizes long-term reward. The vocabulary is compact: agent, environment, state, action, reward, policy, value, and episode. Modern large language models are refined through reinforcement learning from human feedback, which teaches them to prefer responses humans rate higher. Humanoid robots, game-playing systems that beat world champions, and production LLM assistants all share this learning loop.

Part of the free AI Learning Hub by Shahzad Asghar. The whole track takes about 16 minutes and is free to read with no account required.