Integrating Artificial Intelligence in English Speaking Practice: Perceptions, Challenges, and Effectiveness in Vocational Higher Education

Abstract

The progress of computers that can think like humans has opened new chances for learning languages, especially for speaking, which has always been one of the hardest things for students to master. This research aims to examine what teachers and students think about using computer programs powered by AI, such as ChatGPT, Duolingo, Elsa Speak, and Google Gemini, to help them improve their speaking. The way it was done was by using a mix of methods that started with numbers and then moved to words. The part with numbers had surveys for students and teachers, and then the part with words looked closer at what the surveys showed. The data showed that most people thought AI was very useful for practicing speaking, mainly for learning how to say words, how smoothly they spoke, and which words they used, and for giving quick answers that made learning faster. AI is also seen as adaptable because it lets people practice on their own, whenever and wherever they want, and it can get students excited with fun activities and games. But the biggest problems found were not having enough good technology, some switch students not being ready to use technology well, and AI not being able to teach as well as people because it does not understand the situation and cannot fully replace talking to a real person. In conclusion, the research shows that AI could significantly improve the effectiveness of learning to speak. But this can only happen if AI is used alongside teachers who are involved and with the right technology to support it. The research stresses that AI should be seen as a resource that enhances language learning, not as a replacement for teachers. This highlights how important it is to come up with better ways to include AI that fit the situation, are fun, and can keep going in colleges, which should be what future research and putting into practice focus on.