演算法社會的管制——多元管制以及全球行政法的可能

Abstract

演算法在現代生活中對法社會的行動者影響層面十分廣闊，同時左右著個人生活及政治生活。針對演算法的相關議題，國內文獻多關注於如何透過法律建制達到個人資料保護（涵蓋對於大數據科技的討論，以及人工智能的法律議題討論）的目的。而在這些議題的背後，較少針對在這些科技產品的背後扮演著重要角色的演算法技術本身進行法律議題的形塑。演算法既無形體，也並非法社會中的行動者，但其之所以進入到法律世界中，仍有著使用它的個人。 本文藉由：第一，作為人工智慧核心的演算法有哪些特色？在這裡，本文試圖對演算法科技進行淺略的介紹，並討論目前以人工智慧的方式稱呼這些科技是否合適。並試圖找出，演算法作為自動化決策機制之所以會有問題最可能的原因為何。 第二，演算法的大量使用對社會可能有什麼樣的影響？在該章中，將以對個人生活及政治生活的可能影響進行分析。因為這種種影響，認為應該有管制的必要性。 第三，面對演算法社會，有哪些管制的可能性？在這章中，討論了目前主流的以個人資料為中心的管制方式的利弊。接著介紹以演算法為中心的管制方式的可能性。最後，討論以公害為類比的觀點，並引出需要從全球的角度來看待演算法社會的問題。 第四，如何以全球的角度看待演算法社會的成形？在該章中，先討論全球化可能的型態，最後引入全球行政法的觀點，試圖預測將來可能的發展。

In modern society, the use of algorithm has a significant impact, affecting both personal life and political life, on the actors in the society of law. In Taiwan, curent legal studies about algorithm relating subjects were mostly focusing on legal construction of personal information protection law, which including the big data technology, and AI issues. Besides all these issues, algorithm, which is an important factor behind most modern technologiey, is seldom discussed by the legal research society in Taiwan. Although algorithm lacks substance, and is unable to act in the sociey of law, it can enter the world of law by the person using it. This thesis is fousing on: first, what are the characters of algorithms, being the core of artificail intelligence ? In the first part of the thesis, we will do a brief introdution about algorithms, and discuss whether the name “artificial intelligence” suits this art. And we will try to find the most possible reason of why algorithms being automatic decision-maker may cause problems. Second, what is the impacts of the heavy use of algorithm in society ? In this part, we will analyze in aspect of personal life and political life. Because these impact, we may come to a conlusion that regulation may be needed. Third, what kind of regulation is possible, when facing the algorithmic society ? In this part, we will discuss pros and cons of the mainstream regulation which is a personal information-based regulation. Then a possible approach of algorithm-based regulation is introduced. And we will use an analogy of public nuisance to dicuss the regulation of algorithmic society. Therefore we need to see this problem in a global aspect. Forth, How to form a global aspect of algorithmic society ? In this part, we will dicuss possible forms of the globalization of algorthmic society. And we use the aspect of the global administrative law in order to predict the future.