Making AI systems feel guilty

LLMs and GPT systems has amazed everyone with excellent results for question and answering problems being able to pass exams and tests with ease. However we don’t see these systems being conscious and we don’t see them having emotions. This talk will provide some ideas on how create empathic AI systems and how higher order human like emotions like guilt,

Speaker Bio:
Dr. Ronjon Nag, President of R42 Group, Adjunct Professor in Genetics, Stanford School of Medicine.

Additionally, he provides free AI education through the R42 Institute In 2016, he became a Stanford Interdisciplinary Distinguished Careers Institute Fellow and teaches courses on AI, genes, ethics, longevity science, and venture capital.

Ronjon is a Fellow and Trustee of the Institution of Engineering and Technology (IET), a Lifetime Member of the Association for Computing Machinery (ACM) and a Senior Member of the IEEE. He has received the IET Mountbatten Medal, the Verizon Powerful Answers Award ($1 million prize), the 2021 IEEE-Santa Clara Valley Outstanding Engineer Award, and the 2023 COGX AI Lifetime Achievement Award. Ronjon received a PhD from Cambridge, an MS from MIT, a BSc from Birmingham University (UK) and was a Stanford Harkness Post-Doctoral Fellow. He is Chairman of Ecrio (5G systems) and Bounce Imaging (throwable cameras). Ronjon is also part owner of some 100 AI and Biotech start-ups. He is the 2024 Silicon Valley Engineering Council Hall of Fame Inductee.

0:00 Chapter Intro
8:34 Speaker Intro
9:59 Presentation
10:45 Highlights - Who am I
15:54 What is intelligence
19:01 What is artificial intelligence
31:00 Boundaries of Humanity
38:11 Can a machine have a mind
48:08 Emotion Modelling
1:01:25 Computational Emotional Models
1:05:16 Simulating guilt - does it mean it feels guilt?
1:10:08 Chat to me (Soopra URL)
1:10:17 Q&A