All Episodes

Memory and reflection, in brief

Orchestration and shared state, in brief

The capstone, in brief

How an agent fetches its own data, in brief

The bull and the bear, in brief

The risk gate, in brief

The trader, in brief

Why split one AI into many, in brief

AI-authored commits and PRs: brief

Git branches, in brief

Cherry-pick and stash: brief

Commit hygiene, in brief

Merge conflicts: brief

Multi-agent integration patterns: brief

Pull requests, in brief

Rebase, deeper: brief

Releases and tags: brief

Remotes and forks: brief

Git team workflows, in brief

The future of git in an AI world, in brief

Undoing things in git: brief

Why git exists: brief

Worktrees and parallel agents, in brief

Your first repo: brief

API keys and OAuth: brief

CostGuard and privacy: brief

First conversation and model picker: brief

Clawless memory system: brief

AI governance, in brief

AI safety as a field: brief

Beneficial AI and machine ethics, in brief

Multi-agent AI collective action: brief

Complex systems and emergent risk: brief

Four catastrophic AI risks: brief

Monitoring and robustness, in brief

Safety engineering, in brief

The alignment problem: brief

Shipping a Claude application: brief

Subagents and Claude Managed Agents: brief

Deep RL open problems: brief

Exploration strategies: brief

Multi-task RL and meta-RL: brief

Offline RL algorithms: brief

Offline RL: brief

Diffusion models II: brief

Score-based diffusion via SDEs: brief

Four-paradigm landscape, in brief

Agent Skills and Claude Code: brief

Single call to agent loop, in brief

Model Context Protocol, in brief

Prompt caching and context management, in brief

Six effective-agent patterns: brief

Choosing your model and the effort dial, in brief

Server-side tools and built-ins, in brief

Messages API in production: brief

Tool use, in brief

Your first Claude API call: brief

Securing agents, in brief

Multimodal agents in production, in brief

Where multimodal AI is going, in brief

Data filtering and deduplication: brief

LLM data sources, part 1: brief

Language model evaluation: brief

GPUs and TPUs: brief

Inference: brief

Kernels, Triton and XLA: brief

Parallelism, in brief

Post-training, SFT and RLHF: brief

Reasoning and RLVR, in brief

Scaling laws, in brief

Bias-variance tradeoff: brief

Classification metrics: brief

Train, test, cross-validation: brief

PCA, in brief

t-SNE: brief

Recovering 3D vision, in brief

CNN architectures, in brief

Convolution and CNNs: brief

Detection, segmentation, visualization: brief

Diffusion models: brief

GANs and VAEs: brief

Human-centered computer vision, in brief

Linear classifiers, in brief

Loss and optimization: brief

Neural networks and backprop: brief

Self-supervised vision, in brief

Sequence tools for vision: brief

Video understanding: brief

Vision and language: brief

World modeling: brief

Actor-critic methods: brief

Control as inference: brief

DQN (replay, target, double-Q): brief

Imitation learning, in brief

Deep reinforcement learning, in brief

Model-based RL, in brief

Planning with a learned model: brief

Policy gradients (REINFORCE): brief

PPO clipped surrogate: brief

RL fundamentals: brief

RLHF pipeline: brief

Value-based RL, in brief

Variational inference for RL: brief

Autoregressive models: brief

Diffusion models I: brief

Energy-based models, in brief

Evaluating generative models: brief

GANs, the minimax game: brief

Latent variables and the ELBO: brief

Maximum likelihood and the KL view: brief

Normalizing flows: brief

Score matching, in brief

VAE reparameterization trick, in brief

WGAN-GP and Wasserstein loss, in brief

Generative model paradigms: brief

LLM agents: brief

Augmented language models: brief

Industry perspective, in brief

Launch an LLM app, in brief

LLM foundations: brief

LLMOps, in brief

Project walkthrough: brief

Prompt engineering: brief

Training your own LLM, in brief

UX for language user interfaces: brief

LLM landscape in motion: brief

Large multimodal models, in brief

JEPA and world modeling: brief

Multimodal world models for science: brief

Native multimodal intelligence, in brief

Reasoning over multimodal inputs: brief

Transformers for video generation, in brief

Diffusion image generation: brief

Multimodal AI: brief

Function approximation: brief

Markov Decision Processes: brief

Monte Carlo prediction, in brief

Policy gradient and modern RL, in brief

Policy iteration: brief

Q-learning, in brief

Temporal-difference learning, in brief

Value functions and Bellman: brief

Value iteration: brief

Reinforcement learning, in brief

Higher-order derivatives: brief

Taylor series: brief

Attention alternatives and MoE: brief

Counting the cost: brief

From scratch and the tokenizer: brief

The Transformer architecture: brief

Boosting: brief

Decision trees, in brief

Linear regression: brief

Hierarchical clustering: brief

Gradient descent, in brief

k-means clustering, in brief

Logistic regression: brief

Random forests, in brief

Support vector machines: brief

Machine learning, in brief

Why seeing is hard: brief

Backpropagation, in brief

Gradient descent: brief

Neurons and layers, in brief

Neural networks recap: brief

The cost landscape: brief

The whole network as one function: brief

Weights and biases, in brief

Backpropagation: brief

What learning really means, in brief

Build and share a demo, in brief

Curating datasets: brief

Debug your training: brief

Fine-tune a pretrained model: brief

Fine-tuning LLMs, in brief

Reasoning models, in brief

Run a model in a few lines: brief

Share on the Hub: brief

The main NLP tasks: brief

Tokenizers up close: brief

Wrangling data with Datasets: brief

Bayes' theorem: brief

Conditional probability, in brief

Confidence intervals, in brief

Hypothesis testing and p-values: brief

Probability foundations: brief

Random variables and expected value: brief

Sampling and the central limit theorem: brief

Statistics in machine learning, in brief

Summarizing data, in brief

The binomial distribution: brief

Normal distribution: brief

Data distributions and histograms, in brief

Correlation, in brief

Why AI runs on statistics: brief

The chain rule: brief

The essence of calculus: brief

Implicit differentiation: brief

Integration and the fundamental theorem, in brief

Limits and L'Hopital's rule: brief

The power rule from geometry: brief

The product rule, in brief

The derivative as a rate, in brief

Trig derivatives from geometry: brief

Why area equals slope: brief

Why e is special: brief

3D cross product via duality: brief

Stepping up to 3D: brief

Abstract vector spaces, in brief

Change of basis, in brief

Cramer's rule: brief

Cross products: brief

The determinant: brief

Dot products: brief

Eigenvectors and eigenvalues: brief

Matrix inverse and null space: brief

Linear transformations: brief

Matrix multiplication: brief

Matrices between dimensions: brief

Spans and basis, in brief

What vectors are: brief

What transformers do: brief

Agentic RAG, in brief

Building trustworthy agents, in brief

Choosing an agent framework: brief

Giving agents memory: brief

Tool use: brief

Agents that self-check: brief

Multi-agent systems, in brief

Planning: breaking a goal, in brief

Tool-use design pattern, in brief

What makes an AI an 'agent': brief

The handwritten-digit problem: brief

BERT architecture, in brief

BERT pretraining and fine-tuning, in brief

Chain-of-thought prompting, in brief

Agent loops, in brief

Function calling, in brief

How models know word order: brief

Reasoning models, in brief

LLM-as-a-Judge: brief

Few-shot prompting, in brief

Speculative decoding and diffusion LLMs: brief

RLHF and DPO: brief

Transformers beyond text: brief

AI safety threads, in brief

Why benchmarks can mislead, in brief

Why tool-using models fail, in brief

Preferences into reward signals: brief

Pretraining, in brief

Parallelism and Flash Attention: brief

Quantization and mixed precision, in brief

Scaling laws and Chinchilla, in brief

Attention efficiency, in brief

DistilBERT and RoBERTa, in brief

Encoder-decoder, T5, span corruption: brief

How prompting works, in brief

How RAG works, in brief

LayerNorm, pre-norm, RMSNorm: brief

RoPE position embeddings: brief

Instruction tuning (SFT), in brief

How a transformer generates text: brief

Multi-head attention, in brief

The transformer block, in brief

Embeddings: word vectors, in brief

Tokenization, in brief

Self-attention, in brief

AI won't replace you, in brief