JOURNAL OF DIALECTICS OF NATURE
A Comprehensive, Academic Journal of the Philosophy, History, Sociology and Cultural Studies of Science and Technology
Author Information
XU Yingjin
Information
<p>School of Philosophy</p><p>Fudan University</p><p>yjxu@fudan.edu.cn</p>
Research Articles
-
What if the “Chinese Room” is Updated as the “Japanese Room”? On How to Make an Embodied AI System for Natural Language Processing Abstract: In his original version of the “Chinese Room Argument”, John Searle appears to be more interested in refuting the “Strong AI thesis” as a philosophical thesis on the nature of mind rather than revealing some empirical problems embedded in the practices of Natural Language Processing (NLP) . Hence, even if the “Chinese room” could be replaced by a “Japanese room”, the structure of Searle's argument would not be thereby changed. However, Searle's indifference to the difference between natural languages, say, that between Chinese and Japanese, blocks his way to apprehend a critical feature of Japanese, namely, the very relevance of the embodied situations of the speakers to certain expressions. And the ignorance of this feature also makes him ignorant of how challenging it is for mainstream approaches in NLP to do justice to the phenomenal feelings of Japanese speakers, or how difficult it is for a computer program to pass the Turing Test on the linguistic capacities for Japanese. In this sense, to update Searle's original “Chinese room” as a “Japanese room” would greatly facilitate our critical assessment of the mainstream NLP approaches' capacity for handling linguistic problems related to embodied situations of human language speakers, although assessment of this type is not so relevant to the plausibility of the Strong AI thesis as merely a metaphysical issue.Issue:Volume 40, lssue 1, January 2018Page: 8-16
-
How to Make Artificial Intelligence Capable of Speaking Human Language? Some Philosophical Remarks on Natural Language Processing Abstract: There are four salient problems in the current state of Natural Language Processing (NLP): (1) There is no real integration among different sub-areas of NLP; (2) There is no real integration among NLP and other branches in AI; (3) The limited successes of big-date-based NLP approaches are based on the exploitation of human labor; (4) Big-date-based NLP approaches do not perform well when they are employed to handle linguistic data with rhetorical features. The most NLP-related philosophical problems include: (1) Do human language primarily represent the external world or the internal world of the language-users? (2) Are linguistic rules innately fixed in human minds or results of the influence of language-learning environments? (3) What about the relationship between linguistic expressions and human cognitive architecture? (4) In what sense and what degree is it necessary to make language-related cognitive architecture “embodied”? Key Words: Artificial intelligence (AI); Natural language processing (NLP); Cognitive linguistic; Philosophy of language; Big-date; SchemaIssue:Volume 44, lssue 1, January 2022Page: 10-19
-
Reflections on the Empiricist Assumption of Large Language Models: A Study from the Perspective of Miki Kiyoshi’s Theory of Imagination Abstract: From the Japanese philosopher Miki Kiyoshi’s point of view, the empiricist notion of “experience” is too thin to accommodate temporal extension, reference to external objects, as well as the interplay between experience-owing agents and their environments. Once properly modified, the foregoing Miki’s criticism can be easily applied to the still-fashionable approach of Large Language Models (LLM) in AI. Paralleled with the empiricist philosophy, the LLM-builders construct the digital counterpart of the Humean notion of “impressions”, namely, “tokens” by treating the text as tokens. Moreover, similar to the Humean route-map of reconstructing “ideas” from the accumulations of “impressions”, LLM-builders also intend to reconstruct the semantic features of tokens via a proper statistical treatment of them, namely, a treatment routinely under the label of “word embedding”. Nonetheless, since such reductionism-oriented route-map has deliberately bypassed nearly all middle/high-level architecture required for a full-fledged notion of “cognition”, such rout-map could hardly account for why human’s cognitive machine can deliver creative decisions and be competent in counter-factual reasoning even when the size of the training data is much smaller than that is required by an empiricist theory. And the LLM-builders’ incompetence of accounting for all of this in turn philosophically explains the origin of the socalled “machine hallucination”. Key Words: Imagination; Empiricism; Large language models; Tokens; Word-EmbeddingIssue:Volume 47, lssue 8, August 2025Page: 1-9
JOURNAL OF DIALECTICS OF NATURE
- Contact Us
- Address: No.19A Yuquan Road, Beijing, 100049, China
- Phone: +86-10-88256007
- Email:jdn@ucas.ac.cn
Follow Us
© 2014 Copyright of the University of Chinese Academy of Sciences
© 2014 Copyright of the University of Chinese Academy of Sciences
