کاربرد هوش مصنوعی در تصمیم گیری سیاست خارجی؛ ظرفیت ها و چالش ها (مقاله علمی وزارت علوم)
درجه علمی: نشریه علمی (وزارت علوم)
آرشیو
چکیده
پیشرفت ها در زمینه هوش مصنوعی عده ای را به این باور رسانده است که در آینده ای نه چندان دور نه تنها تصمیمات سیاسی بلکه تصدی مناصب سیاسی هم به هوش مصنوعی واگذار شود. به همین دلیل این پژوهش درصدد است با اتکا به روش ترکیبی یا آمیخته، به این پرسش پاسخ دهد: هوش مصنوعی در تصمیم گیری سیاست خارجی چه کاربردی دارد و ظرفیت ها و چالش های کاربرد آن چیست؟ فرضیه پژوهش این است که در وضعیت فعلی، هوش مصنوعی قادر به تصمیم گیری مستقل در سیاست خارجی نیست و در عوض در سیاست خارجی در قالب سیستم های پشتیبان و مکمل یا مشاور تصمیم گیری کاربرد دارد. به بیان دیگر یافته ها نشان می دهند که علی رغم اینکه سیستم های هوش مصنوعی در ارائه پشتیبانی تصمیم گیری به سیاست گذاران امیدوارکننده بوده اند، اما هنوز قادر به جایگزینی قضاوت و شهود انسانی در تصمیم گیری سیاست خارجی نیستند. در نتیجه بهتر است که هوش مصنوعی را به عنوان یک ابزار مکمل به جای یک تصمیم گیرنده مستقل در نظر بگیریم. همچنین اگرچه در بسیاری از موارد، تصمیم گیرندگان انسانی به کمک هوش مصنوعی می توانند تصمیمات بهتر و قابل اعتمادتری بگیرند اما چالش های آن نیز مواردی چون عدم دسترسی به داده های با کیفیت و سوگیری های الگوریتمی، آسیب پذیری در برابر اطلاعات جعلی دشمن، ملاحظات اخلاقی و غیره است.Applications of Artificial Intelligence in Foreign Policy Decision-Making: Capacities and Challenges
Introduction AI refers to systems that exhibit intelligent behavior by analyzing their environment and taking actions, with some degree of autonomy, to achieve specific goals. In recent years, AI has also made its way into politics. In an increasingly interconnected world, the emergence and rapid development of AI has become a pressing issue in foreign policy decision-making. The exponential growth of data and the emergence of advanced AI technologies have opened new opportunities to enhance human decision-making capabilities. The integration of AI technologies into foreign policy decision-making processes has the potential to transform the way countries shape their strategies and navigate complex geopolitical challenges. The present study aimed to examine the role of AI in foreign policy decision-making, as well as the capacities and challenges lie therein.Literature ReviewSeveral studies have dealt with to the use of AI in foreign policy decision-making. For instance, Berkoff (1997) explored the potential of AI to enhance individual and organizational decision-making capabilities, offering suggestions to the AI engineering community on what policymakers need to improve these processes. Putri et al. (2020) used AI to simulate ASEAN negotiations. They clarified how AI decision-making systems can help international relations experts and learners understand the decision-making process and government foreign policy strategies during diplomatic negotiations. Scott et al. (2018) tried to provide a basis for planning a foreign policy strategy that effectively responds to the emerging power of AI in international affairs. In the Iranian context, there is a notable lack of research regarding AI and its impact or application in foreign policy or international relations.Materials and MethodsThe study employed a mixed methods research design to analyze over 100 sources. Lawshe’s content validity ratio (CVR) method was used to validate the research hypothesis. Moreover, a panel of 12 experts, including academics and AI specialists, evaluated the hypothesis. With a CVR value exceeding 0.56, the hypothesis was confirmed, which showed the reliability of the research findings.Results and DiscussionThe applications of AI in foreign policy decision-making can be classified under two main types: 1) AI independent decision-making as in the virtual politician or AI politician and 2) AI as a support and supplement to decision-making in foreign policy. Currently, the concept of the AI politician is still at the theoretical level and has not been implemented in practice. Despite the efforts to promote the AI politician, these systems face many challenges and limitations. AI cannot understand complex foreign policy dynamics, cultural nuances, historical context, and diplomatic subtleties. Additionally, AI cannot evaluate and measure moral considerations or take responsibility for decisions. It is also unable to explain unexpected events, sudden policy changes, or emerging trends in world affairs. As a result, until these challenges are resolved, it is unlikely that humans will allow AI systems to make independent decisions.AI tools and methods (e.g., neural networks, expert systems, fuzzy logic, evolutionary computing, natural language processing algorithms, and computational argumentation technology) can be used to support and supplement decision-making in foreign policy. Artificial neural networks can be used to predict conflicts and armed confrontations. A concrete example can be found in Olaide and Ojo (2021) which developed a model to predict conflicts in Nigeria. Expert systems can aid in diplomatic negotiations by creating scenarios. An example is the simulation of foreign policy decision-making in ASEAN negotiations. Fuzzy logic can also be helpful as seen in Sanjian’s attempt to use the fuzzy set theory for modeling the decision-making process in the U.S. arms transfers (Sanjian, 1988). Genetic algorithms are the most widely used evolutionary computing methods in decision-making, which can shorten decision-making time and provide better decision-making plans for foreign policy decision-makers. In addition, natural language processing algorithms can reduce language barriers between countries, allowing diplomats and policymakers to communicate more easily with foreign governments and embassies. Moreover, special systems like the Project Debater can debate complex issues with humans, helping people make persuasive arguments and informed decisions.Using AI in foreign policy decision-making presents both capacities and challenges. AI offers several capacities to support decision-making in foreign policy. By helping decision-makers analyze abundant and diverse foreign policy data, AI enhances the ability to better assess decision risks and reduce errors in decision-making. This leads to more accurate predictions of international events and speeds up the decision-making process. Moreover, AI can reduce the influence of human biases, emotions, and interests, ultimately improving the quality of foreign policy decisions and enhancing the overall diplomacy process. Regarding the challenges, it is important to note that if the data inputted to AI is biased or incomplete, it can replicate and perpetuate human prejudices such as racism and sexism. The improper or unethical use of AI also poses a significant challenge. Additionally, AI systems lack the capacity for innovation and creativity and do not understand the value of human life or the consequences of military operations. Another challenge is verifying the performance of AI systems. The black box problem also presents a challenge, as the decisions made by AI cannot be easily explained by humans. Overall, considering both the capacities and challenges, researchers recommend that AI should support human strategic decision-making, with final decisions always being made by people in positions of authority.ConclusionThe present study examined various applications of AI in foreign policy decision-making. The discussion concluded that the concept of the AI politician is not yet practical since AI lacks the capability to fully comprehend and navigate the complexities of foreign policy, thus making the emergence of such a phenomenon unlikely. In other words, AI cannot replace humans in making independent decisions within the realm of foreign policy. However, AI is effectively used today as a support and supplementary tool, serving roles in analysis and prediction to aid decision-making in foreign policy.
تبلیغات
