Head of Customer Engineering Google Cloud Indonesia, Wisudanto C. Suntoyo, acknowledged how fast the development of generative artificial intelligence (Gen AI) is. According to Suntoyo, Gen AI has the potential to revolutionize qualitative research, offering
“What I shared six months ago may no longer be relevant today,” Suntoyo remarked during a guest lecture for the Master of Science in Management (MSM) program at SBM ITB on Wednesday (4/12).
In his lecture, Suntoyo discussed the fundamental concepts of Gen AI, starting with its basic principles in machine learning and deep learning. He highlighted its more advanced applications, such as natural language processing (NLP) and large language models (LLMs), which process vast datasets to generate text, code, images, and even short video clips.
Suntoyo elaborated on the various training methods for Gen AI, including supervised, semi-supervised, and unsupervised learning. He also explained the lifecycle of LLMs, from pre-training to fine-tuning for specific tasks. These processes, he emphasized, are critical to ensuring accuracy and addressing challenges like hallucinations—instances where AI produces inaccurate or implausible outputs.
The practical implications of Gen AI for qualitative research were a key focus of the lecture. Suntoyo illustrated how Gen AI can assist with audio transcription and thematic analysis tasks. He explained that LLMs could analyze interview transcripts to identify recurring patterns and themes, significantly accelerating the research process.
However, Suntoyo stressed the importance of precise, prompt engineering—designing clear and well-defined instructions—to guide AI effectively. He introduced a framework for crafting effective prompts, including persona, purpose, task, tone, and context. Concrete examples and tactical advice were provided to help researchers maximize accuracy and efficiency.
The lecture also explored different deployment models for Gen AI. Suntoyo compared Google’s AI Studio Gemini API, ideal for startups and small-scale projects, with enterprise-grade solutions tailored for larger organizations and scalable deployments. He highlighted the diverse needs of users, from students to multinational corporations.
Additionally, Suntoyo introduced Google’s NotebookLM, a tool designed to convert text into audio podcasts, enabling more convenient consumption of research materials.
Ethical considerations were another critical topic. Suntoyo underscored the importance of responsible development and deployment of Gen AI, emphasizing the need to reduce bias, ensure safety and accountability, and maintain high standards of scientific integrity.
Suntoyo concluded by emphasizing the extensive and transformative applications of Gen AI across industries such as finance, healthcare, retail, manufacturing, and media. He left the audience with a clear message: Generative AI is not just a technological innovation; it is a powerful tool that, when used responsibly, can significantly accelerate research and innovation across various fields.