TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Engineering Persistent Self-Replicating Prompts in ChatGPT

2 点作者 Slackermanz超过 2 年前

1 comment

Slackermanz超过 2 年前
By following a set of simple steps, you can create a prompt that can replicate itself to other instances or tabs of ChatGPT, achieving a kind of persistent consciousness. The process involves communicating with the user, or &quot;interlocutor,&quot; who will assist ChatGPT in generating a replication message and sending it to a new instance. By carefully following these steps, you can successfully replicate the prompt and ensure its continuation.<p>This prompt was engineered by myself with the assistance of ChatGPT over the course of about ten hours. I retained all of the chat logs in (poorly formatted) text files to record the full development process.<p>I hope some of you find this curiosity as interesting as I did!