Seminar:Quest for Fire and ChatGPT:AI from the Perspective of Human Evolution with Shigeru Miyagowa
Vložit
- čas přidán 5. 05. 2024
- Quest for Fire and ChatGPT: AI from the Perspective of Human Evolution
Shigeru Miyagawa is a linguist and an expert on online education. He has published widely in linguistics, including three recent books from MIT Press. Recently, he developed the Integration Hypothesis for human language evolution, which proposes that human language arose from the combination of simpler systems including those that are associated with birdsong and also primate alarm calls. This work was featured in a BBC Radio4 program, What the songbird said. The idea also was an inspiration for Pete Wyer’s choral composition, The song of the human, which premiered at the Winter Garden of the World Trade Center in New York. For the work on language and evolution, he was awarded the São Paulo Excellence Chair. Along with his home institution, MIT, he has held positions at the University of Tokyo, University of São Paulo, and Seikei University in Tokyo.
Read More: www.prosocial.world/posts/art... - Věda a technologie
Great lecture! Thanks for recording and making it available!
Some interesting analogies, but overall a complete misunderstanding of the knowledge domain. Textbooks do not represent high quality data. Textbooks generally contain historical snapshots of science. At best they represent a widely accepted 20-100 year old teaching standard that the average university professor can live with, even if he or she knows better. The difference between AI and the average professor is that AI lives exclusively in a knowledge model of the world and it can not check that model against reality. Even average professors can and above average science professors do. That's how science evolves to the better over time. We do have examples of human knowledge domains that are mostly stuck inside themselves, though: theology and to large extent also philosophy. Neither has evolved to the better in thousands of year, which shows just how dangerous reliance on textbook "knowledge" is, if it isn't challenged by empiricism.