AI人工智慧對公共政策在政策倫理的對話初探

The Dialogue between Artificial Intelligence and Public Policy upon Policy Ethics

自從1960年代開啟了人工智慧（Artificial Intelligence）的魔法門，AI人工智慧一直是一個充滿衝擊性的話題。近年來隨著相關研究的發展與運用科技的成熟，AI人工智慧不僅是近年來最受到矚目與討論的科技話題，也是最引人擔憂與煩惱的公共議題。隨著AI人工智慧相關的研究與技術紛紛進入多門學科，很多困擾人類已久的問題，如重大疾病，可能都能夠迎刃而解。但另方面，AI的發展帶來的社會衝擊，可能會徹底瓦解人類最基礎的社會、經濟、文化與生活的結構。因此，對於AI的發展，學術社群與實務領域的先驅者均充滿著憂喜參半的情緒。
表面上看起來，AI人工智慧的蓬勃發展，與相關科技運用的進步，似乎會為人類未來的生活帶來無盡的便利，這種對於AI人工智慧充滿樂觀的想法，是促使AI人工智慧突飛猛進的原動力，所以，某些學者認為相關機構如學術界與政府不應該限制AI的發展，反而應該運用不需許可創新（Permissionless Innovation）的概念，鼓勵AI的各種應用發展，鼓勵百花齊放，直到出現明顯而立即的危機方休。但是在另一方面，有識之士與相關利益團體，卻對AI相關科技的發展提出嚴重警告，告誡人類如果過分依賴AI人工智慧等的發明，不僅在可預見的未來，人類社會的秩序與結構立刻會面臨挑戰，如勞工政策與社會安全政策可能會面臨全面崩潰的情形，而產生更為嚴重的社會衝突；甚或可能有一天人類全體會受到AI人工智慧體的反噬與控制，導致全體人類淪為AI人工智慧體的奴隸、甚或滅亡。這種對於AI人工智慧的幻想、期許、讚嘆、與恐懼，以及發展AI相關科技對於倫理道德的挑戰。
本文認為，未來的AI發展無論其重點為何，都必須嚴格遵守醫學科技發展的倫理道德要求，經過一套嚴格的審查機制，才能控制脫序的潛在威脅。

Ever since Dr. He Jiankui revealed his success in delivering the first pair of gene-editing-to-fight-AIDs babies in 2018, almost all scientists blamed such a reckless endeavor. Yet such an advancement in technology also realizes the prolong fear for humans that technology will someday prevails over humanity. To make the matter worse, another cloud looming the horizon that might offended the humanity even worse is the development of artificial intelligence (AI in short).
Ever since the hatchling days of AI in the 1960’s, the advancement in the field is full of controversy. Along with the latest development in technology and application, the AI has become the hottest, as well as the most worrisome topic in scientific world and public policy. The AI has made many improvement possible in many fields like medical science and industrial automation, yet it also posts serious threats and ethic challenges to society, culture, economy, and humanity.
Some scientists with optimal attitude argue that there shall be a“Permissionless Innovation”on the development of the AI, while others worries about the infringement of AI into many sensitive policy areas with uncontrollable consequences which result in the total destruction of humanity.
This paper argues that in order for the AI to serve and help humans to progress without possible heavy costs, a safety mechanism regarding the ethics of the any research should be introduced and established by duplicating the stringent examination process as currently utilized by the medical society when it comes to clinical human trials.