watch list · curated

Learn fast. Build faster.

Hand-picked tutorials worth your time before the next launch window — each with an honest note on why it earns a slot. Videos play right on the page; creators keep full credit and every view.

[1hr Talk] Intro to Large Language Models — video thumbnail

beginner-friendly

[1hr Talk] Intro to Large Language Models

by Andrej Karpathy · plays via YouTube — every view counts for the creator

The single best hour you can spend before an AI hackathon. Karpathy (OpenAI founding member, ex-Tesla AI lead) explains what an LLM actually is, why it behaves the way it does, and where the attack surface lives — no jargon tax. Watch at 1.25x; the security segment at the end is directly stealable for demo-day framing.

Let's build GPT: from scratch, in code, spelled out — video thumbnail

intermediate · code-along

Let's build GPT: from scratch, in code, spelled out

by Andrej Karpathy · plays via YouTube — every view counts for the creator

Two hours, one file, a working GPT. This is the code-along that turns “I use the API” people into “I understand the model” people. You will not build this at a hackathon — but judges can smell who has done it. Do it once, slowly, before your next event.

you need to learn MCP RIGHT NOW!! (Model Context Protocol) — video thumbnail

beginner · hands-on

you need to learn MCP RIGHT NOW!! (Model Context Protocol)

by NetworkChuck · plays via YouTube — every view counts for the creator

The fastest on-ramp to MCP — the protocol behind the GSA MCP & AI Agent Hackathon. Chuck wires LLMs to real tools via Docker, then builds a custom MCP server from scratch — Kali Linux one included. High caffeine, zero gatekeeping. Skip to 12:31 if you only want the build.

📼 this list stays alive

New tutorials get added as they earn it — same rule as the events: verified, credited, worth your weekend. Made one that belongs here? Get in touch.