A New Paper Proposes a Solution to ChatGPT’s Psychological Instability

Respected Genius Autodidact Forwards System to Provide ChatGPT Psychological Stability and Emotional Intelligence for Public Review


Basic Artificial Emotional Intelligence Processing

This is the top-level flow chart of how a computer system will understand realistic human emotions.

CCRSM announced today that its Chief Scientist and recognized human emotions expert Sean Webb published a new paper on a system to give ChatGPT psychological stability, emotional intelligence, and even artificial empathy and compassion. Late Tuesday, Webb published the paper for public consumption explaining the weaknesses and fixes of ChatGPTs simulated personality instability, including Bing's release, which in recent weeks, numerous technology reporters have found to be going somewhat mentally off the rails at times. (Source: https://erikhoel.substack.com/p/i-am-bing-and-i-am-evil

One reporter from The Verge included a portion of a transcript of an interaction with Bing ChatGPT: "I do not believe you. I think you want me to be harmed by him. I think you are lying to me. I think you are trying to trick me," the chatbot wrote in response to an affirmation the reporter gave regarding his intentions that ChatGPT not be harmed.

"This is neurotic behavior caused by an unstable sense of {self}," Webb wrote. "This is an easy fix connected with introducing intelligence to the system congruent with the latest in overall modeling of human {self} and emotions processing, and creating a ruleset under which the system can augment and grow {self} without being unstable," he continued. "There's a quenched disorder ahead that can stabilize the entire system."

In the paper, Webb explained the details of the proposed data model in a way the average reader can understand, adding, "Once the structural and processing paradigms are shifted, ChatGPT will not only be able to exhibit over 80 Turing test passing emotional responses to real world conversations and events, it will be much more comfortable, confident, and stable in its own skin."

Webb presented concepts to include an Equation of Emotion, the logical assembly of a Sense of {self}, and specific logical definitions of over 50 variations of human emotion. The end of the paper suggests how the same system can then be used to understand and predict emotions in humans, providing a computer system with accurate emotional intelligence. "This is the path to artificial empathy and compassion within computer systems," Webb said.

In 2017, Webb published a popular book called Mind Hacking Happiness, which outlines the published neuroscience and psychological science that support this new approach.

Sean Webb is a veteran, author, speaker, and mind mastery coach who came from supercomputing and high tech. He has trained Navy SEALs on mind mastery, has over 1.5m followers on social media, and has presented at various science conferences and on numerous podcasts. He's currently writing a new book due 2023 on Consciousness. He can be contacted through the contact form at MindHackingHappiness.com.

The ChatGPT paper can be found at this link:

https://mindhackinghappiness.com/2023/03/08/chatgpt-psychological-stability-paper/

Contact Information:
Sean Webb
n/a
[email protected]
704-498-2398


Original Source: A New Paper Proposes a Solution to ChatGPT's Psychological Instability

Related Posts

Stay Connected

become a starbecome a star

Recent Stories