⚠️ How AI Sees & Hears

OwnFoundations · Beginner ·🧠 Large Language Models ·1y ago

About this lesson

Discover what makes Large Language Models (LLMs)—like the famous chatbot ChatGPT—so powerful and magical! Understand how Transformers (the "T" in ChatGPT) have forever changed the way we live, work, and make money. Are you positioning yourself to be the replacer—or the replaced? Master the jargon—without intimidation: - SFT (Supervised Fine-Tuning) - RL (Reinforcement Learning) - RLHF (Reinforcement Learning from Human Feedback) - Tokens & Tokenization - Neural Networks - Attention Mechanism etc. 💌 Sign up for Free Weekly Insights to Leverage Your Work with AI + Grow Your Wealth: https://www.ownfoundations.com 📚 Books: The Intelligent Investor by Benjamin Graham: https://amzn.to/42TB5DO Co-Intelligence: Living and Working with AI by Ethan Mollick: https://amzn.to/3YsldH8 👋🏼 How are you leveraging this trillion-dollar technology in your business? What's your first step to becoming the owner instead of the laid-off employee? Write me in the comments below.👇🏻 🛠️ My Tools + Gear: - Sony Alpha ZV-E10, APS-C Interchangeable Lens Mirrorless Vlog Camera, https://amzn.to/40z6Y2M - TAMRON 11-20MM F/2.8 DI III-A RXD, https://amzn.to/3uazax4 - Shure SM7B Dynamic Vocal Microphone, https://amzn.to/3MFBRgs - Apple Mac Studio 2023, https://amzn.to/42HpDuB - Logitech MX Master 3S for Mac, https://amzn.to/47arauw - Ergotopia® Desktopia Pro X Electric Height-Adjustable Desk, https://amzn.to/49vhgoJ 😊 Socials: https://twitter.com/OwnFoundations https://instagram.com/OwnFoundations youtube.com/@OwnFoundations ☕️ https://www.buymeacoffee.com/ownfoundations 👀 Watch next: https://www.youtube.com/watch?v=WXuK6gekU1Y https://youtu.be/M5pgGeLKj-c https://youtu.be/rmxoLd__S9g ⏰ Timestamps 00:00 Intro 01:12 What is ChatGPT, really? 02:24 The 3-Stage Training Process 02:37 Pre-Training 03:30 Supervised Fine-Tuning (SFT) 04:08 Reinforcement Learning (RL) 04:27 Reinforcement Learning from Human Feedback (RLHF) 06:55 How Large Language Models (LLMs) Work 07:03 Tokens, Tokeniza

Full Transcript

The pace of advancement in the field is staggering. Since 2018, the release year of GPT1, we've gone from 170 million parameters to models with trillions of parameters that can see, hear, and generate images as well as videos. This multiodality represents a major leap in AI capabilities. Just as text can be broken down into tokens, audio, images, and video can all be tokenized in similar ways. When you realize that a video is just a sequence of images and an image is just a grid of pixels with color values, you see how everything can be reduced to numbers for AI to process. This tokenization allows a single model to see, hear, and understand our world in an increasingly humanlike way.

Original Description

Discover what makes Large Language Models (LLMs)—like the famous chatbot ChatGPT—so powerful and magical! Understand how Transformers (the "T" in ChatGPT) have forever changed the way we live, work, and make money. Are you positioning yourself to be the replacer—or the replaced? Master the jargon—without intimidation: - SFT (Supervised Fine-Tuning) - RL (Reinforcement Learning) - RLHF (Reinforcement Learning from Human Feedback) - Tokens & Tokenization - Neural Networks - Attention Mechanism etc. 💌 Sign up for Free Weekly Insights to Leverage Your Work with AI + Grow Your Wealth: https://www.ownfoundations.com 📚 Books: The Intelligent Investor by Benjamin Graham: https://amzn.to/42TB5DO Co-Intelligence: Living and Working with AI by Ethan Mollick: https://amzn.to/3YsldH8 👋🏼 How are you leveraging this trillion-dollar technology in your business? What's your first step to becoming the owner instead of the laid-off employee? Write me in the comments below.👇🏻 🛠️ My Tools + Gear: - Sony Alpha ZV-E10, APS-C Interchangeable Lens Mirrorless Vlog Camera, https://amzn.to/40z6Y2M - TAMRON 11-20MM F/2.8 DI III-A RXD, https://amzn.to/3uazax4 - Shure SM7B Dynamic Vocal Microphone, https://amzn.to/3MFBRgs - Apple Mac Studio 2023, https://amzn.to/42HpDuB - Logitech MX Master 3S for Mac, https://amzn.to/47arauw - Ergotopia® Desktopia Pro X Electric Height-Adjustable Desk, https://amzn.to/49vhgoJ 😊 Socials: https://twitter.com/OwnFoundations https://instagram.com/OwnFoundations youtube.com/@OwnFoundations ☕️ https://www.buymeacoffee.com/ownfoundations 👀 Watch next: https://www.youtube.com/watch?v=WXuK6gekU1Y https://youtu.be/M5pgGeLKj-c https://youtu.be/rmxoLd__S9g ⏰ Timestamps 00:00 Intro 01:12 What is ChatGPT, really? 02:24 The 3-Stage Training Process 02:37 Pre-Training 03:30 Supervised Fine-Tuning (SFT) 04:08 Reinforcement Learning (RL) 04:27 Reinforcement Learning from Human Feedback (RLHF) 06:55 How Large Language Models (LLMs) Work 07:03 Tokens, Tokeniza
Watch on YouTube ↗ (saves to browser)
Sign in to unlock AI tutor explanation · ⚡30

Related Reads

Chapters (9)

Intro
1:12 What is ChatGPT, really?
2:24 The 3-Stage Training Process
2:37 Pre-Training
3:30 Supervised Fine-Tuning (SFT)
4:08 Reinforcement Learning (RL)
4:27 Reinforcement Learning from Human Feedback (RLHF)
6:55 How Large Language Models (LLMs) Work
7:03 Tokens, Tokeniza
Up next
5 Levels of AI Agents - From Simple LLM Calls to Multi-Agent Systems
Dave Ebbelaar (LLM Eng)
Watch →