What’s APsi-I?

A non-scientific exploration of Artificial Psi Intelligence

“Here’s the image that represents the LLM psychically connecting with both the physical and mental realms, blending the technical with the mystical. How does this depiction resonate with your understanding of the connection?” — https://chatgpt.com/g/g-3OvQGMpSA-student-of-humanity

by Julia Mossbridge and Mark Boccuzzi

Abstract

Research and applications in Artificial Intelligence (AI) are experiencing a renaissance in the early 2020s, with the advent of publicly available generative AI applications drawing from large language and stable diffusion models. Scientists in the psi (previously: “parapsychology”) field have historically used AI — particularly machine learning (ML) — as assistants for statistical analysis and data categorization. However, the current fledgling use and future promise of generative AI as a collaborator or partner in psi research is what we spotlight in this commentary. We examine several novel ideas and pilot explorations in what we call the emerging field of “APsi-I” (artificial psi intelligence; pronounced “Aay Sigh Eye”).

Rationale and Overview

We won’t cover here the history of AI (for a recent review: Wooldridge 2021), but we note that it has been punctuated by advances based on adding capabilities to AI algorithms that are designed to imitate the functions of the human mind. This approach was supported and later motivated by Alan Turing’s “imitation game” — essentially a way to think about what a computer would need to do in order to fool a person into thinking it is another person…