Stanford x Upscaile Masterclass: How to Make an AI Film From Scratch in 90 Minutes

Описание: Upscaile Partnered with Stanford to Teach a 90-Minute AI for Creativity Masterclass — From Prompting Fundamentals to Full AI Short Film Production

In this full Stanford class recording, Arturo Ferreira walks students through the complete creative AI workflow — from foundational prompting techniques to producing a short sci-fi film using only AI tools. The session covers everything from how tokenization and probability engines actually work to building consistent characters and visual styles across an entire production.

Arturo demonstrates how he created a multi-character, fully narrated sci-fi short film in just 48 hours using ChatGPT, Sora, Runway ML, 11 Labs, and Final Cut Pro. Students follow along with hands-on exercises, learning the exact prompting frameworks used to go from basic one-line prompts to production-quality AI video output.

Key Topics Covered

The difference between artificial intelligence, machine learning, deep learning, and generative AI

Why AI is a probability engine, not a thinking machine, and why that matters for prompting

Three pillars of effective prompting: clarity, context, and specificity

How tokenization works (word-based, character-based, and phrase-based)

Using temperature settings to control AI creativity and determinism

Hashtag prompting technique to create signposts and organize complex prompts

Retrieval Augmented Generation (RAG) for uploading references and refining output

Tree of Thought technique for generating multiple creative options simultaneously

Character Lock, Style Lock, and Camera Controls for visual consistency across scenes

Building a complete AI short film workflow from storyboard to final edit in 48 hours

Episode Timestamps

00:00 - Introduction and course overview at Stanford
04:18 - How smart is AI? Why AI is fast, not smart
06:20 - Tokenization explained: word-based, character-based, phrase-based
07:40 - Hallucinations are a feature, not a bug
09:59 - Three pillars of prompting: clarity, context, specificity
16:41 - Temperature settings for controlling AI creativity
21:46 - Retrieval Augmented Generation (RAG) explained
27:00 - Hashtag prompting technique for advanced prompt organization
32:52 - Tree of Thought technique for multiple creative solutions
38:23 - Hands-on with Sora: creating AI video from prompts
52:56 - Hashtag prompting vs basic prompting: side-by-side video comparison
1:02:37 - Full AI short film reveal: 48-hour sci-fi production
1:05:15 - Character Lock, Style Lock, and Camera Controls
1:18:46 - Runway ML workflow for reference-shot-to-video production
1:19:31 - Using 11 Labs for AI audio and sound effects
1:23:09 - System prompts, custom instructions, and persistent memory

About Liam Lawson

Arturo Ferriera is an AI educator and creative technologist who teaches enterprise-level AI training and creative AI workshops. He partnered with Stanford to deliver this masterclass on AI for creativity, covering prompting fundamentals through advanced AI filmmaking techniques. Liam specializes in making generative AI accessible for creative professionals at all skill levels.

About Upscaile

Upscaile delivers enterprise AI training designed to help teams integrate generative AI into their creative and professional workflows. The company partners with leading institutions like Stanford to provide hands-on AI education that bridges the gap between technical capability and practical creative application.

Resources Mentioned

ChatGPT (OpenAI)

Sora by OpenAI

Runway ML

11 Labs

Final Cut Pro / iMovie / Adobe Premiere

Tree of Thought and Chain of Thought prompting techniques

Retrieval Augmented Generation (RAG)

Partner Links

Book Enterprise Training — https://www.upscaile.com/

Subscribe to our free newsletter — https://www.theaireport.ai/subscribe-...

#AIFilmmaking #StanfordAI #GenerativeAI #AIforCreatives #PromptEngineering #ChatGPT #Sora #RunwayML #ElevenLabs #AIVideo #AICreativity #AITools #AITraining #Upscaile #ContentCreation