Using AI Chat Bots to Assist Callers During Emergencies
Ehtesham Ahmed Shariff 1, Suriyakala Perumal Chandran 2 Large language models (LLMs), such as ChatGPT, have attracted widespread public interest due to their perceived benefits and concerns. Although the potential contribution of LLMs to healthcare has been widely discussed, few studies have carefully examined the application of LLMs, including their use in the personal agreement process. This study evaluated the effectiveness of LL.M students, ChatGPT 3.5, ChatGPT 4.0, and Gemini in creating information on the anesthesia process. It is a pity that despite many attempts no LLM has been able to produce legal documents and the reserve for risk prevention, disclosure, and planning remains inadequate. Therefore, the LLM now clarifies the limitations in practical use, particularly in changing the patient's risk classification by agreement. Interactive agents, including chatbots and interactive voice interaction (IVR), have made significant progress and promise further to advance human-machine interaction, especially in emergency procedures. This article describes a two-manned emergency solution for short-term emergencies using Free Pass IVR and the Bot press chatbot platform, a multi-purpose solution. The solution was evaluated in Romania to demonstrate quality of service (QoS) and effectiveness of service management. Additionally, proof of concept (PoC) was assessed using real data, focusing on current and valid emergency calls in Romania. According to the feasibility study, the PoC is ready to integrate into Romania's emergency; The IVR phone is best suited to the requirements of the country's system.