Source: Zhongshi Electronic News
Author: Lai Xiangwei
AI is popular.For a while, Wei was learned, and almost nothing without any artificial intelligence (Artificial Intelligence).Among the areas of AI, machine learning has attracted much attention. As for further and further deep learning of feature engineering, it can automatically learn and find derivative variables that humans may not be noticed. ThereforeThe prediction effect is often better.
Al can be used to distinguish fake messages and can also be used to make fake messages.However, the current research shows that if you want to distinguish the authenticity of the information from the form, you can only cope with the false information manufactured by AL; for artificially manufactured false information, it is not easy to identify it directly by AL.This technical confrontation should continue to improve both sides. For example, when AI recognizes false information, it is mainly based on text. AI manufacturing fake messages has been improved to deep pseudo -films.
It is undeniable that AI is an effective tool for anti -fake messages, but AI must not be regarded as the only criterion for judgment with the final judgment.Because from the perspective of the theory of freedom of speech, the earliest Milton emphasized that it was judged by human rationality to judge the truth and right or wrong, and the truth may be discovered for a long time, and even the Mill who will never be discovered;It is the right to emphasize the speculation and choice of truth. By the 20th century, the "conceptual market" exposition proposed by the US Supreme Court Justice Homs was even more brilliant.
In recent years, some scholars have started from the perspective of empiricalism. To verify whether the market is in the market failure because of the proliferation of false information.At the same time, the idea of identifying fake messages through AI is even more popular.The combination of the two is almost possible to infer that AI is a tool to better distinguish the authenticity of information than human opinion.
This is actually completely wrong.First of all, the inheritance of the knowledge theory of freedom of speech is not an empiricalism, but a pragmatism, that is, believe that such claims are beneficial to the progress of human society.Because of this, even if many people don't believe that "the truth is more and more clear," it always supports freedom of speech, which reflects the spirit of pragmatism.
Secondly, to explore whether the market failure has occurred in the market, we must first know where the conceptual market is?This market does not exist in public opinion space such as the Internet, but that it exists in everyone's own mind, that is, the independent judgment ability of each citizen, which is a market market.In this regard, the discussion of others, scholars' research, and AI identification are just references. They must not replace each citizen's judgment of true and false or wrong.Once AI replaces human judgment, it is deprived of human rational rights.
AI is a great tool, but it must not rely on AI to fight fake messages, but to rely on the rational speculation of citizens.All AI should worry about not having robot rule, but that the totalitarian rule of the AI will be controlled.
The author is a professor at the Department of Radio and Television of National Taiwan University of Art
