Generative AI is Impressive but deceptive 生成式人工智能令人赞叹，但极具误导性

Image generated by Microsoft Bing Image Creator at https://www.bing.com/create

In the realm of text generation, text-based conversations, and voice dialogues, the current prevailing Artificial Intelligence technology relies on Large Language Models, reminiscent of the late Taiwanese writer Li Ao. These AI models possess vast knowledge and an exceptional memory, but they also tend to indulge in exaggeration. While they can serve as insightful advisors, their reliability as strategic or execution partners may not always be guaranteed.

On the other hand, in the domain of image, video, and voice generation, contemporary Artificial Intelligence excels at fabrication, simulation, and counterfeiting. It can adeptly create deceptive and realistic content, making it a powerful tool for deceiving audiences and generating false impressions, particularly when used without tangible supporting resources. For individuals lacking access to concrete resources, this undoubtedly becomes a convenient instrument for misleading others.

总体而言，在文本生成，文字对话和语音对谈方面，现在以大型语言模型为主流的人工智能，就像是已故台湾作家李敖一样，很有学识，博闻强记，但喜欢胡吹；可以是很有洞见的智囊，但作为策略或执行伙伴，则不一定靠谱。

在图片、视频和语音生成方面，当下的人工智能擅长虚构、模拟和伪造，能够以假乱真、鱼目混珠，对于那些没有足够实体资源的人来说，这无疑是一种混淆视听和虚张声势的利器。