A conference room was filled with software engineers, researchers, and entrepreneurs as they gathered to learn about the latest developments in generative artificial intelligence (AI). Hosted by Galaxy Software Services (GSS), a leading technology company in Taiwan, the event focused on the origins of AI and its applications in chatbots and computer vision.
GSS, which services a wide range of industries from finance to healthcare, is at the forefront of the AI revolution in Taiwan. As part of their commitment to staying ahead in this field, GSS organizes an annual forum to discuss the opportunities and challenges in AI.
The forum began with Perry Chang, Chairman of GSS, sharing his journey in the field of AI. Chang initially ventured into AI to develop a simple Chinese grammar and spell-check function. After facing difficulties, he eventually succeeded in creating a natural language processing (NLP) system, which marked the beginning of GSS’s involvement in AI.
Chang highlighted the significant advancements in NLP, from basic optical character recognition (OCR) to the development of chatbots that can respond to preset cues. He emphasized the importance of large language models (LLM) in enhancing the accuracy and quality of chatbots.
According to Chang, Taiwan needs its own LLM that comprehends traditional Chinese characters and reflects the nation’s cultural heritage. He also emphasized the need for data confidentiality and cost-effectiveness in creating a native LLM.
The forum also featured Jane Hsu, Chair of the National Taiwan University Computer Science and Information Engineering Department, who discussed the importance of developing a secure database for LLM. Hsu introduced the Trustworthy AI Dialogue Engine (TAIDE), a project dedicated to compiling Taiwan-specific information, including language, culture, and history.
Hsu explained that while classical machine learning requires vast amounts of labeled data, the use of generative pre-trained transformers (GPT) has revolutionized the AI landscape. However, the increasing computational demands and costs associated with AI development pose challenges for Taiwan, urging the government to provide increased funding.
The development of databases like TAIDE and the enhancement of domestic AI capabilities are crucial in nurturing Taiwanese talent in AI. Hsu stressed the significance of maintaining data quality, preventing biases, and ensuring trustworthiness and cybersecurity.
Despite the challenges, the convergence of chatbots and AI, exemplified by the creation of ChatGPT, presents enormous opportunities for GSS. The company can leverage its extensive experience in implementing AI solutions to meet market demands.
In conclusion, the forum shed light on the need for Taiwan to develop its own native AI database for chatbots. The speakers emphasized the importance of data integrity, confidentiality, and cultural relevance. With advancements in AI and the concerted efforts of industry and academia, Taiwan can establish itself as a key player in the global AI landscape.
