Rachel Rishita | Polygence
Symposium presenter banner

Symposium

Of Rising Scholars

Fall 2025

Rachel will be presenting at The Symposium of Rising Scholars on Saturday, September 27th! To attend the event and see Rachel's presentation.

Go to Polygence Scholars page
Rachel Rishita's cover illustration
Polygence Scholar2025
Rachel Rishita's profile

Rachel Rishita

Class of 2026Millcreek, WA

About

Hello! My name is Rachel Rishita and I am very excited to learn more about the different implications of the negative effects of Artificial Intelligence.

Projects

  • "From Fiction to Framework: Disentangling Instrumental Convergence from Practical AI Risk" with mentor Aditya (May 15, 2025)

Project Portfolio

From Fiction to Framework: Disentangling Instrumental Convergence from Practical AI Risk

Started Dec. 4, 2024

Abstract or project description

This paper debunks the hypothetical situation presented by instrumental convergence -- the belief that Artificial Intelligence will develop sub-goals which will make it indifferent to its previous programming and unstoppable against humans -- by stating how AI lacks intrinsic desires, self-preservation instincts, and agency unless explicitly programmed with such tendencies. In other words, lacking these characteristics which would motivate it to develop a selfish character is unrealistic and a product of anthropomorphism, which wrongly associates AI with human attributes that make instrumental convergence possible. Having such associations with AI is dangerous as it may lead to its mischaracterization, overestimation, and lack of accountability for the people that developed its bugs and errors. Overall, instrumental convergence is unrealistic and provides additional barriers to AI Alignment with human values and understanding of its capabilities.