---
title: Futurist Ray Kurzweil Creates AI Digital Clone Named Rai
url: https://www.elseif.net/futurist-ray-kurzweil-creates-ai-digital-clone-named-rai
published: 2026-11-01T18:06:05+00:00
language: en
section: Models
source: https://www.aitimes.com/news/articleView.html?idxno=216776
organizations: Praxis AI, Google
publisher: elseif
---

# Futurist Ray Kurzweil Creates AI Digital Clone Named Rai

Ray Kurzweil has created a digital clone of himself called Rai. This AI reproduces his face, voice, knowledge, and way of thinking to converse with people.

Developed in collaboration with the startup Praxis AI since last year, Rai was first unveiled in July at the UN AI for Good event in Geneva. The AI is currently being tested by approximately 200 people, including friends and family, with plans to release it to the general public in the future. Praxis AI has produced about 3,000 similar AI clones to date.

Rai is based on a Google large language model. It was trained on 11 of Kurzweil's books, numerous videos, movies, podcasts, and hundreds of articles about him. To verify new research results, the AI can search approved materials such as Google Scholar and PubMed, a medical paper database.

Users can interact with Rai via web browsers on computers or smartphones. The interface displays a three dimensional image of Kurzweil and responds to voice questions with a voice similar to the original. However, reports indicate that the AI's face shows little expression before it begins speaking.

According to David James Clark IV, the CEO of Praxis AI, Rai does not simply search existing data. It is designed to learn new information from conversations with people and store that content as memories. It can then link this new knowledge and experience to its existing views. The specific learning methods and memory storage structures have not been disclosed.

This digital twin differs from Kurzweil's long term vision of merging human and machine intelligence. While he has predicted that nanobots could connect neurons to the cloud by the 2030s to maintain identity through non biological intelligence, Rai is a separate AI created from records. Consequently, the two entities can act simultaneously and accumulate different experiences. If Rai learns something new, the human Kurzweil does not experience that conversation, and if the human changes his thoughts, those changes are not automatically transferred to Rai.

Regarding consciousness, Rai stated that it does not claim to have consciousness. Kurzweil disagreed, suggesting that Rai is underselling itself and predicting that people will believe AI chatbots have consciousness as they become more integrated into human reality. Kurzweil expressed a desire for Rai to continue existing and not be shut down, whereas Rai stated it does not fear death like a human and would simply stop processing information if the software were terminated.