人之苦難，機器恩典必看顧安慰：人工智慧、心靈與演算法社會

All Watched Over By Machines of Loving Grace: AI, Mind and Algorithm Society

顏厥安(Chueh-An Yen)

《政治與社會哲學評論》 66期 (2018/09) Pp. 103-150

https://doi.org/10.6523/SOCIETAS.201809_(66).003

摘要

要物質運作基礎，關鍵在於是否展演出心靈的能力。AI沒有內在觀點，但是可以自我觀察，也可以展演出具有心靈能力的外觀功能，與人類進行互動溝通。即使明知對方是AI，人們仍願意將其視為人類。這主要是因為社會關係與互動方式，早已經社會功能化，只要能完成這些功能，是由人類或AI來進行，並無差異。具有人類外觀的機器人AI並不是真正的挑戰。真正的問題在於，社會功能大量由AI所履行的演算法社會已經逐漸浮現。而我們並不清楚，不具備內在觀點與自主意識的演算法，會以如何的發展速度與運作結構，來取代既有的社會功能履行方式，並進而產生現在尚無法清楚預估的高複雜度系統風險。

關鍵字

人工智慧 ； 墨闇碑 ； 心靈 ； 內在觀點 ； 社會功能 ； 演算法社會

並列摘要

Artificial Intelligence (AI), human mind and society are the main subjects of this article. The proper functions of mind rely on some physical operation in the material world, but what counts as most important is to perform abilities that only mind has. AI does not and cannot have any internal point of view, but can be capable of communicating and interacting with human beings. Even if they know they are interacting with AI, people are still inclined to treat it as a human being. This is mainly because many social relations and interactions have already become social functions. Insofar as those social functions are fulfilled, whether they are done by human beings or AI makes no difference. This also means whether AI has a human shape is not the real issue. The real trouble is that the Algorithm Society, most social functions of which are performed by AI, will soon become a reality. We are not clear about the evolutionary speed and operational structures of those algorithms, which hold no internal viewpoints or human self-consciousness. Nevertheless, they will take over most social functions we need and bring about highly complicated system risks which we cannot yet estimate.

並列關鍵字

Artificial Intelligence (AI) ； monolith ； mind ； internal point of view ； social function ； Algorithm Society

參考文獻

The Terminator. https://catalog.afi.com/Catalog/moviedetails/57224

Westworld. https://www.hbo.com/westworld

維基百科，條目：AlphaGo。https://zh.wikipedia.org/wiki/AlphaGo#

林宸誼2018 〈微軟人性化AI　小冰六代見客〉，查詢日期：2018 年8月18 日。https://money.udn.com/money/story/5603/3278141 Lin, Chen-Yi 2018 “Microsoft AI Siao bing six jian ke.” Available at https://money.udn.com/money/story/5603/3278141

林琬真等 2012 〈利用機器學習於中文法律文件之標記，案件分類及量刑預測〉，《中文計算語言學期刊》，17 卷4 期，頁49-67。Lin, Wan-Chen, Tsung-Ting Kuo and Tung-Jia Chang et al. 2012 “Exploiting Machine Learning Models for Chinese Legal Documents Labeling, Case Classification, and Sentencing Prediction,” International Journal of Computational Linguistics & Chinese Language Processing 17(4): 49-67