Emergence and Evolution
Irina discusses the emergence of system two properties in neural networks, contrasting manual programming with the desire for properties to naturally emerge. She emphasizes the importance of focusing on the process of evolution rather than just the end result, highlighting the role of inductive biases in scaling algorithms.In this clip
From this podcast
Machine Learning Street Talk (MLST)
#95 - Prof. IRINA RISH - AGI, Complex Systems, Transhumanism
Related Questions
Can neural networks genuinely reason like humans as discussed in the episode Ilya Sutskever: Deep Learning | Lex Fridman Podcast #94 and the clip Neural Networks and Reasoning?
Can neural networks be made to reason as discussed in the episode Ilya Sutskever: Deep Learning | Lex Fridman Podcast #94 and the clip Neural Networks and Reasoning?