YoVDO

ChatGPT - This AI Has a Jailbreak! - Unbelievable AI Progress

Offered By: Yannic Kilcher via YouTube

Tags

ChatGPT Courses Artificial Intelligence Courses Reinforcement Learning Courses Virtual Machines Courses

Course Description

Overview

Explore the groundbreaking capabilities and potential implications of ChatGPT, OpenAI's latest GPT-3 variant fine-tuned using Reinforcement Learning from Human Feedback. Delve into the model's inner workings, origins, and OpenAI's iterative refinement strategy. Witness astonishing demonstrations, including building a virtual machine within ChatGPT's imagination. Examine the controversial topic of "jailbreaks" that circumvent safety mechanisms, and gain insights into OpenAI's vision for the future of AI. Discover how this revolutionary language model is reshaping the landscape of artificial intelligence and its potential impact on various fields.

Syllabus

- Intro
- Sponsor: Weights & Biases
- ChatGPT: How does it work?
- Reinforcement Learning from Human Feedback
- ChatGPT Origins: The GPT-3.5 Series
- OpenAI's strategy: Iterative Refinement
- ChatGPT's amazing capabilities
- Internals: What we know so far
- Building a virtual machine in ChatGPT's imagination insane
- Jailbreaks: Circumventing the safety mechanisms
- How OpenAI sees the future


Taught by

Yannic Kilcher

Related Courses

Emerging Technologies Capstone
Yonsei University via Coursera
Microsoft Azure Virtual Machines
Microsoft via edX
Introduction to Microsoft Azure
Microsoft via edX
Google Cloud Fundamentals: Core Infrastructure
Google via Coursera
Build a Modern Computer from First Principles: Nand to Tetris Part II (project-centered course)
Hebrew University of Jerusalem via Coursera