| 摘要: | 本研究旨在探討因高齡化社會出現而對人工智慧產生的需求，特別是ChatGPT聊天機器人和人形機器人在高齡化長期照顧領域中的應用，以及這些應用如何影響照顧者及其相關工作照顧者的關係及品質。
ChatGPT透過人工智慧根據輸入文字做出相對應回覆的一種文本生成技術。在技術初期，因AI資料庫有限等等限制，常出現答非所問或回答不完全準確的情形。但隨著更多訓練數據、技術日趨成熟再加上對大量文本的分析，ChatGPT能給予越來越精準的答覆。它在發布後引起廣泛的關注和討論，例如機器倫理、隱私保護和誤導性資訊等問題。就像大家常用的iPhone中的Siri，也是一種透過AI給予回答的工具。
本研究對象為年滿18歲且有長期照顧經驗或相關工作照顧者，通過問卷收集數據，以分析各個面向，包括信任和滿意度，對於照顧服務的態度忠誠度和行為忠誠度的影響。
研究發現，人工智慧在長期照顧服務中的運用在照顧者及其相關工作者之間存在顯著的正向關係。為了增強照顧者使用機器人的信心，建議推廣機器人教育培訓。接下來，通過建立產業、政府與學術界三方合作的組織，讓照顧者看到使用機器人為改善高齡者照顧的未來。最後針對機器人在實際應用中可能遇到的盲點與困難，進行針對性改善和人性化設計，以增強照顧者對機器人實用性和重要性的認知。研究亦探討了照顧服務提供方式和技術專門性對於這種關係的調節作用，結果顯示，在技術專門性較高的情況下，信任與忠誠度的正向關係會降低，而滿意度與忠誠度的正向關係則會增強。這一發現為提高高齡者對長期照顧服務的忠誠度提供了策略性建議，並揭示了不同照顧服務方式對於高齡者忠誠度的影響。
This study aims to explore the application of artificial intelligence, specifically ChatGPT chatbots and humanoid robots, in the field of long-term care for the aging population, as well as how these applications affect the relationships and quality of care among caregivers and their related work caregivers.
ChatGPT is an AI chatbot program developed based on the GPT model, fully named as Chat Generative Pre-trained Transformer, developed by OpenAI, focusing on generating conversational agents. ChatGPT operates by making judgments and generat-ing responses based on the text input through AI, a text generation technology. Initially, due to limitations in AI's database and computational capabilities, it might produce irrel-evant or inaccurately complete answers. However, with more training data, improved model architectures, and maturing technology, coupled with extensive text analysis, ChatGPT can provide increasingly accurate responses. It supports multiple languages, catering to users from different linguistic backgrounds, and has garnered widespread attention and discussion upon release, sparking many issues such as machine ethics, pri-vacy protection, and misinformation, similar to how Siri on the iPhone functions as an AI response tool. To enhance safety and effectiveness, OpenAI continues to improve the model and take measures to address these issues.
The study targets individuals over 18 years old with experience in long-term care-giving or related work, collecting data through surveys to analyze various aspects, in-cluding trust and satisfaction, attitude loyalty, and behavioral loyalty (such as reselection and recommendation) towards care services.
Findings indicate a significant positive relationship between the use of AI in long-term care services and its caregivers and related workers. To boost caregivers' confi-dence in using robots, the study suggests promoting robot education and training. Fur-thermore, by establishing collaborative organizations among industries, governments, and academia, support from long-term care-related organizations can demonstrate to caregivers the future of improving elderly care with robots. Lastly, targeted improve-ments and human-centric designs should address potential blind spots and difficulties in robots' practical application to enhance caregivers' recognition of robots' utility and sig-nificance. Additionally, the study examines the moderating role of service delivery methods and technical specialization on this relationship, finding that higher technical specialization diminishes the positive relationship between trust and loyalty, while the positive relationship between satisfaction and loyalty is strengthened. These findings offer strategic suggestions for enhancing the elderly's loyalty to long-term care services and reveal the impact of different care service methods on the elderly's loyalty. |
