意识研究视域下中国机器翻译产业发展的探源性分析

How Consciousness Impacts China’s Machine Translation Industry

近十年随着人工智能技术的迅速发展,人工智能在人类生产生活领域的应用也逐渐深入,机器翻译也从最初的基于规则、统计和实例的翻译系统转向了人工智能翻译系统,机器翻译依靠"大数据+深度学习"取得了突破性进展。GOOGLE最新发布的人工智能GNMT系统能够以句子为单位准确、自然地翻译文章,部分译文质量已经接近人工翻译。当AlphaGo战胜了人类围棋冠军的时候,机器翻译是否会取代人工翻译也成为了人们热议的话题。文章试图从意识研究角度分析人脑翻译机制的过程以及人工翻译区别于机器翻译的原因,从而给中国的机器翻译产业发展趋势提供新的启示。该研究发现,人类意识贯穿着翻译活动的整个过程,而机器无法实现对这些主观性的思维过程进行全过程模拟,因而人脑与机器在翻译活动中可以做到优势互补和有机结合,从而实现高质量高效率的翻译目标。

Over the past ten years,artificial intelligence (AI) technologies have been flourishing and extensively applied in people’s life and work.Machine translation (MT) has achieved an unprecedented success by drawing upon the combination of big data and deep learning,differing from the previous rule-based,statistics-based and example-based systems.In fact,Google’s newly-issued GNMT system has proved to be capable of translating texts sentence by sentence in an almost accurate and natural fashion as compared to human translation.The controversial debate over the question whether MT will replace human translation or not is heated after AlphaGo beat the human champion in a Weiqi game.This paper endeavors to analyze the human brain translation mechanism and the differences between MT and human translation in light of consciousness studies and thus inspire the future MT industry development trends in China.The current research finds out that consciousness dominates the entire process of translation activities but MT is insufficient in terms of simulating these subjective experiences while translating.The interactive cooperation between human and machines in translating practices enables high-quality and efficient translation outcomes.