言++ 以环境叙事的方法将AI学习人类语言的过程具象化为一场探索之旅。
Yan + + visualizes the process of AI learning human language into an exploration journey by means of environmental narration.
现代互联网语境下语言已发生了显著变化,语言及语言背后的语义急剧弱化甚至与其原本所表达的情感信息严重不匹配,在此情况下大量高强度的“新语言”增生。
正如今天“哈哈”两字在互联网的语境下早已徒留敷衍的内涵,想要让对方信服我们是真的高兴至少要七个哈起步。然而从什么时候开始我们不用这样极端的语言就无法表达自己真正的情感了?当我们的语言习惯最终都变成单一定性的网络语言,建立在其上的思维方式又将会发展成什么状态?
言++ 以环境叙事的方法将AI学习人类语言的过程具象化为一场探索之旅。懵懂 的AI男孩将深入人类的语言世界,以孩童般的视角体味语言通货膨胀的发展,玩家将深入这场交互旅行感受不断弱化的语义和极端的语言如何一步步哺育AI男孩并最终将他塑造为一个“完整的人类”。
以专辑形式记录下MAMO教育孩子过程中所生成的音乐,专辑CD的音乐、封面、文本均完全由AI生成。有趣的是,当下人工智能生成音乐的训练模型中,英文歌曲占绝对数量,因此当我
们以中文儿歌为样本生成时不难发现它似乎把基调从中国变得更“西方化”了
这是否显示了西方学者在训练AI生成音乐时携带着“偏见”?在
如MAMO一样的AI技术的帮助下,发达世界的影响最终会否对
其他地区的文化与教育造成侵害?
Record the music generated in the process of MAMO's education of children in the form of an album. The com-position samples come from the well-known Chinese child-ren's songs. The music, cover and text of the album CD are completely generated by AI. What’s interesting is that in the current AI-generated music training model, English son-gs account for an absolute number, so when we gene-rate Chinese children’s songs as a sample, it is not difficult to find that it seems to have changed the atmosphere from China to more “Westernized”
Does this show that Western scholars carry "prejudices" when training AI to generate m-usic? With the help of AI technology like MAMO, will the influence of the developed world eventually cause harm to the culture and education of underdeveloped regions?
当下越来越多的学者探讨如何训练AI、如何“喂养”高质量的数据集,但却少有人探讨AI对人的“反哺”。
如果有一天人与机器的角色互换由AI来教育(训练)人类,它真的是客观且理性的吗?它又是否会在教育孩子的同时携带偏见?
At present, more and more scholars are discussing how to train AI and how to "feed" high-quality data sets, but few people discuss the "feedback" of AI to people. If one day the roles of humans and machines are exchanged and AI is used to educate (train) humans, is it really objective and rational? Does it carry prejudice while educating children?
Design a free website - Click now