Judging machines: philosophical aspects of deep learning
- PDF / 349,439 Bytes
- 21 Pages / 439.37 x 666.142 pts Page_size
- 52 Downloads / 177 Views
Judging machines: philosophical aspects of deep learning Arno Schubbach1 Received: 19 July 2018 / Accepted: 26 February 2019 © Springer Nature B.V. 2019
Abstract Although machine learning has been successful in recent years and is increasingly being deployed in the sciences, enterprises or administrations, it has rarely been discussed in philosophy beyond the philosophy of mathematics and machine learning. The present contribution addresses the resulting lack of conceptual tools for an epistemological discussion of machine learning by conceiving of deep learning networks as ‘judging machines’ and using the Kantian analysis of judgments for specifying the type of judgment they are capable of. At the center of the argument is the fact that the functionality of deep learning networks is established by training and cannot be explained and justified by reference to a predefined rule-based procedure. Instead, the computational process of a deep learning network is barely explainable and needs further justification, as is shown in reference to the current research literature. Thus, it requires a new form of justification, that is to be specified with the help of Kant’s epistemology. Keywords Deep learning · Machine learning · Artificial intelligence · Algorithm · Computation · Judgment · Explanation · Justification · Kant
1 Introduction The recent and astounding boom of artificial intelligence reached the general public at least since AlphaGo triumphed over one of the strongest Go players, Lee Sedol, in 2016. Today, it seems almost inevitable to stumble upon some eye-catching news about artificial intelligence on a daily basis. Success messages by researchers or sensational announcements by start-up companies are promising to improve medical diagnostics, to prevent crimes through predictive analytics or to convict offenders with the help of facial recognition, while yet others aspire to automate translations or journalist writing, assist the car drivers or even substitute them, and so on and so forth. It seems
B 1
Arno Schubbach [email protected] Department of Humanities, Social and Political Sciences, ETH Zurich, Zurich, Switzerland
123
Synthese
to be a common presumption that artificial intelligence, together with big data, makes one of the most important technologies shaping our future. Therefore, it comes as no surprise that the political representatives see their nation states competing for a key technology and try to outdo each other with research programs. Of course, these prospects produce just as much fear and anxiety as they nurture hopes and utopian desires. In short, our situation can be summed up as follows: “Every day we read that digital computers play chess, translate languages, recognize patterns, and will soon be able to take over our jobs.” (Dreyfus 1992, p. 79) Yet, this description was originally published in 1972 by Hubert Dreyfus, one of the most prominent critics of the first heyday of artificial intelligence research in the 1960 and 1970s that has never lived up to its promi
Data Loading...
