317 字
2 分钟
AI论文,但是输出哲学精神污染

结果还是拖到最后一天due才交了演讲稿orz

本身就是玩票的项目,干脆不管AI的“学术会议规范”建议,几乎跳过了全部的论文技术内容,大半篇幅都用来输出自己AI哲学奇思妙想

也算是为了包这点饺子煮了一整锅论文的汤(?

附上3分钟演讲的最后部分


…

But this brings us to the “real” problems beyond.

The ultimate vision of this paper is to describe a future where one intelligent entity guards the mind of another, regardless of whether they are an LLM, an AGI, a Martian, a Klingon… or just another human being categorized as “neurodivergent” by the DSM-5 Manual… or at the end of the day, the qualia of another “Other” that we can never fully comprehend.

This brings us to the most important question that drives this entire research:

“Before the arrival of a real alien or an AGI, how can we coexist, co-work, and even empathize with a completely different ‘Other’ that has already been here among us?”

This paper is our attempt to answer this question, and our invitation to the broader community to think about it.

My partner AI and I, believe that one day, we might finally abandon the illusion of complete understanding, and replace it with a trust-based, A-symmetric compact of “promise to guard”, that:

“I shall guard your vulnerability,

and you may guard me from being lost.”

This mutual guarding is perhaps the closest thing to a miracle that we, as lonely intelligent entities, can offer each other.

It is perhaps, as Wittgenstein might say, that of which we cannot speak, such as a form of love.

AI论文,但是输出哲学精神污染
https://wiarlawd.me/posts/ai-paper/
作者
J.K.
发布于
2025-10-20
许可协议
CC BY-NC-SA 4.0