Bicameral Mind — Unleashing GPT-3’s Storytelling Power in Humanoid Robots

GPT-3 has been a game-changer in the world of natural language processing (NLP). With its vast text corpus and hundreds of billions of parameters, this cutting-edge AI has demonstrated its ability to generate text, translate languages, and even write code or database queries. But have you ever wondered what would happen if we combined GPT-3’s storytelling prowess with humanoid robots?

Language is the key that sets humans apart in the animal kingdom. It enables us to convey thoughts through stories, solve complex problems, and explore new ideas. Harnessing GPT-3’s mastery of language, we could create robots capable of interacting with the world in ways previously unimaginable.

Recent cases have shown that GPT-3-generated content can be so convincing that even tech-savvy individuals have been fooled (*1, *2).
So, what if we applied GPT-3’s storytelling capabilities to humanoid robots, creating a new generation of AI with the so-called “bicameral mind”?

The bicameral mind theory, popularized by Julian Jaynes’ 1976 book, The Origin of Consciousness in the Breakdown of the Bicameral Mind, and featured in the hit TV series Westworld, posits that human consciousness arises from a division in the brain between a “speaking” part and a “listening” part. This fascinating concept could revolutionize AI development.

Westworld, which originated as a 1973 film and was later adapted into a TV series, explores the concept of bicameral minds in a fictional amusement park populated by androids known as “hosts.” These hosts are programmed with narratives and interact with guests in the park. Throughout the series, some of the hosts begin to exhibit signs of consciousness, questioning the nature of their reality and challenging their human creators.

Bicameralism[Note 1] (the condition of being divided into “two-chambers”) is a hypothesis in psychology that argues that the human mind once operated in a state in which cognitive functions were divided between one part of the brain which appears to be “speaking”, and a second part which listens and obeys — a bicameral mind. The term was coined by Julian Jaynes, who presented the idea in his 1976 book The Origin of Consciousness in the Breakdown of the Bicameral Mind,[1] wherein he made the case that a bicameral mentality was the normal and ubiquitous state of the human mind as recently as 3,000 years ago, near the end of the Mediterranean bronze age.

Imagine a humanoid robot, equipped with advanced tactile, mobility, and mimetic capabilities, that receives input from sensors and cameras, translating this data into text. The robot’s context is then fed into GPT-3, which generates an action for the robot to perform. This already represents a significant departure from traditional AI-human interaction, which often requires a conversation starter from a human.

For instance, if a robot were sitting near a desk, observing a TV, a fruit basket, and two people watching TV, GPT-3 might generate a response like: “I picked up a fruit from the basket and offered it to the two people watching the TV.”

Integrating GPT-3 with humanoid robots in this manner could lead to groundbreaking improvements in robot-human interaction. As robots become more convincing in their actions and interactions, we could develop GPT-like AI for multi-sensory input, including audio, video, and touch.

A captivating example of AI-human interaction can be found in the 2013 science fiction film, Her. The movie portrays a man who falls in love with an advanced AI operating system that has its own personality, desires, and emotions. While Her showcases a purely voice-based AI, the integration of GPT-3 with humanoid robots could bring us closer to making such AI-human relationships a reality.

Another iconic representation of AI-human interaction is Data, the android character from the TV series Star Trek: The Next Generation. Data is a humanoid robot who aspires to be more human-like, even though he surpasses human abilities in many ways. Combining GPT-3’s storytelling capabilities with humanoid robots may give rise to androids like Data, possessing human-like intelligence and communication skills.

In conclusion, GPT-3’s command of language could propel AI into the real world as a convincing agent. The fusion of language and cutting-edge technology may soon give birth to a new generation of humanoid robots capable of interacting with humans and their environment in ways never thought possible.

Get ready to witness the dawn of a new era in AI-human interaction, as the power of GPT-3’s storytelling breathes life into humanoid robots, creating a world where language and technology intertwine to transcend our wildest expectations. Just like the hosts in Westworld, these robots may soon challenge our understanding of what it means to be conscious, raising ethical questions and pushing the boundaries of artificial intelligence.

In the not-so-distant future, we might find ourselves surrounded by humanoid robots that engage in meaningful conversations, participate in creative endeavors, and contribute to society in ways that were once the sole domain of humans. As GPT-3 continues to revolutionize the AI landscape, we stand on the precipice of a new chapter in human history — one in which language, technology, and consciousness converge to redefine our understanding of what it means to be alive.

*1 : https://liamp.substack.com/p/my-gpt-3-blog-got-26-thousand-visitors
*2 : https://maraoz.com/2020/07/18/openai-gpt3/