Recursive Self-Improvement
Jürgen delves into the concept of recursive self-improvement, tracing back to his 1987 diploma thesis. Tim explores the discourse on AI existential risks, focusing on the idea of improving learning algorithms recursively. The discussion highlights the allure and challenges of machines learning to enhance their own learning processes.In this clip
From this podcast
Machine Learning Street Talk (MLST)
Prof. Jürgen Schmidhuber - FATHER OF AI ON ITS DANGERS
Related Questions
How is AI getting up to date and constantly learning about itself?
Is meta-learning a form of intelligence as discussed in the episode Pieter Abbeel: Deep Reinforcement Learning | Lex Fridman Podcast #10 and the clip Hierarchical Learning Insights?
Tell me more about incremental improvement in the development of AI web agents