Nature Podcast

Audio long read: How does ChatGPT ‘think’? Psychology and neuroscience crack open AI large language models

5 snips
May 24, 2024
Scientists are using psychology and neuroscience to understand how AI 'thinks'. The podcast explores the challenges of deciphering AI, the importance of explainable AI for safety, and manipulating AI models. It also delves into the inner workings of chatbots and AI models at a neuron level.
Ask episode
Chapters
Transcript
Episode notes
1
Introduction
00:00 • 3min
2
Deciphering the Enigmatic Nature of AI Through Explainable AI
02:34 • 3min
3
Understanding Large Language Models
05:31 • 8min
4
Unpacking AI Models and Behavioral Manipulation
13:28 • 5min
5
Exploring the Inner Workings of Chatbots and AI Models
18:31 • 2min