On September 14, 2024, the Surabaya Integrated Science and Technology Institute (ISTTS) collaborated with the Amazon Web Services (AWS) community to host the "Indonesia Gen-AI Tour 2024." This event provided a comprehensive overview of Artificial Intelligence (AI), particularly Generative AI and Large Language Models (LLM), which are poised to revolutionize industries in the future.

The event kicked off with a warm welcome from Ir. F.X. Ferdinandus, M.T., Vice Rector of ISTTS, who emphasized the significance of AI in our daily lives. Subsequently, Dimas Khosyi Triangga, the head of the AWS user group Surabaya, introduced the community and outlined upcoming events. Notably, Dimas also announced the Party Rock Hackathon, running until September 30, 2024, with exciting prizes such as a HUAWEI MatePad SE 11" Tablet, a HUAWEI Display 23.8 inch Monitor, and a HUAWEI Smartband 9 (Couple).

The first session was delivered by Esther Irawati, an Associate Professor at ISTTS, who presented on "Generative AI in 2024." Esther explained how Generative AI is revolutionizing digital content creation, allowing users to generate images, videos, and more using simple text prompts. As an example, she demonstrated how AI can create a GIF of a cat dancing the Jaipong, a task previously requiring advanced graphic tools.

She also highlighted the rapid development of Large Language Models (LLM) such as Claude, Cohere, and Stable Diffusion, which enable the creation of high-quality digital content at a lower cost. Esther emphasized the importance of selecting AI tools that align with specific needs and budgets, as well as leveraging cloud infrastructure like Amazon Web Services to effectively develop AI applications.

In the second session, Rio Astamal, a Developer Advocate at AWS, introduced the concept of creating an AI Photobooth using Stable Diffusion. He explained how Stable Diffusion works by adding noise to store image data and then reducing noise to create clearer images when given a prompt. Through a demonstration, he showcased how users can upload their selfies and generate creative images using the Stable Diffusion API.

Dewi Oktaviani, a Data Engineer, presented the third session, focusing on Retrieval Augmented Generation (RAG). RAG enables AI to answer questions based on internal documents by storing data in vector form. Dewi demonstrated how this technology can help companies access and manage their internal data more efficiently.

The Party Rock Hackathon challenge, led by Rio Astamal, was another highlight of the event. Participants were tasked with creating AI applications without coding. A showcased demo could detect positive, negative, and neutral responses from given feedback and automatically summarize the results. This opened up opportunities for participants to engage in AI innovation even without deep programming skills.

Joan Santoso, in the fourth session, discussed how LLM Evaluation can be tested using the Street Fighter 3 game. He demonstrated how AI makes decisions based on character positions within the game and how LLM evaluation is performed in terms of inference speed and performance consistency.

Finally, Mochammad Syafii emphasized the importance of Prompt Engineering in generating contextually relevant AI outputs. He explained techniques such as one-shot and few-shot prompting to improve the quality of AI responses. Syafii also provided practical tips for optimizing AI performance through parameter tuning.

The event concluded with the announcement of winners for the Party Rock Hackathon, with the top three participants receiving awards for their innovative AI creations. One notable innovation was "Ultrafin," an application for managing personal finances by considering the user's current financial condition. Additionally, lucky participants received attractive prizes through a spinning wheel. A workshop led by Rio Astamal on using Amazon Bedrock served as the final session, giving participants the opportunity to experiment with the latest AI technology introduced.

The Indonesia Gen-AI Tour 2024 not only provided insights into the future of AI technology but also demonstrated how AI can be accessible to everyone, both with and without coding skills, paving the way for innovation across various industries.