Toàn cảnh Workshop A.I Thực Chiến 2025 – Nơi hội tụ tri thức và đam mê công nghệ! | VTV24

Key Concepts

• AI thực chiến (Practical AI Competition): A competition focused on applying AI to real-world problems, beyond just coding.
• AI Intelligence: Enhancing the intelligence and accuracy of AI models (e.g., GPT, Grok, Gemini, Llama).
• AI Scaling: Optimizing AI solutions for cost-effectiveness and resource efficiency, enabling deployment on less powerful or cheaper infrastructure.
• Strategic Technologies: Government-identified key technological areas for national development, with AI being a primary focus.
• Niche Market: A specialized segment of a larger market where a company can establish a competitive advantage.
• Minimum Viable Product (MVP): The most basic version of a product that can be released to customers to gather feedback.

Workshop Overview and Objectives

The YouTube video transcript details a specialized online workshop held on October 13th as part of the "AI thực chiến" (Practical AI Competition). This event served as a knowledge forum where participants received practical insights and perspectives from leading experts. The workshop aimed to clarify the competition's objectives, emphasizing the application of AI in real-world scenarios rather than solely focusing on coding proficiency. The problems presented were designed to simulate complex practical issues, going beyond typical academic exercises. This includes optimizing model performance on limited resources, handling volatile and noisy real-world data, and integrating AI solutions into existing production systems.

Expert Panel and Their Contributions

The workshop featured a distinguished panel of experts:

• Dr. Đinh Ngọc Thạnh: Representative of the technology committee for the "AI thực chiến" competition.
• Professor Hồ Tú Bảo: Jury member from the Institute for Advanced Study of Mathematics.
• Professor Trần Anh Đức: Jury member.
• Practical AI Experts: Hoàng Quan Khải, Ngô Hiếu, and Nguyễn Hoài Nam.

Dr. Đinh Ngọc Thạnh highlighted that the competition would expose participants to new facets of AI they might not have encountered before. Professor Hồ Tú Bảo drew inspiration from a national steering committee meeting on strategic technologies, noting that AI is the foremost and most comprehensive of the 11 strategic technologies identified by the state. He emphasized the vast career opportunities in AI due to the abundance of data and AI's core function of analyzing and extracting insights from this data.

Professor Trần Anh Đức's Insights on AI Entrepreneurship

A significant portion of the workshop was dedicated to the insights of Professor Trần Anh Đức, who shared his extensive experience in AI research and commercialization in Vietnam and the US. He offered a candid perspective on the challenges of AI startups, stressing the fierce competition, especially against well-resourced large companies.

Key Arguments and Supporting Evidence:

• Beyond a Good Model: Professor Đức asserted that a technically sound AI model is insufficient for success. It must be complemented by a clear business strategy, the ability to secure funding, and, most importantly, an AI product that genuinely solves a market problem.
• AI as a Tool: His overarching advice was to view AI as a tool for problem-solving, not as the ultimate goal of a project.
• Niche Market Strategy: To survive in the startup environment, teams should identify and focus on a niche market where they can establish a distinct competitive advantage.
• Embracing Failure and Agility: Professor Đức underscored the importance of a mindset that accepts failure quickly and possesses the flexibility to pivot, which are essential skills for any technology founder.

Understanding AI Problem Types: Intelligence vs. Scaling

Professor Trần Anh Đức also delved into the fundamental approaches to developing AI products, categorizing them into two main directions:

1. AI Intelligence: This direction focuses on enhancing the intelligence and accuracy of AI models. Examples include the continuous upgrades of core models by platforms like ChatGPT, Grok, Gemini, and Llama to improve their cognitive capabilities. This is the aspect most participants tend to focus on.
2. AI Scaling: This direction, which Professor Đức noted is often overlooked by Vietnamese participants, addresses the significant cost and resource requirements of AI development. AI Scaling aims to design computational solutions at the infrastructure level. The goal is to enable the same AI model to run on cheaper and more accessible infrastructure, rather than solely relying on expensive, high-performance servers (e.g., from Nvidia). This could involve deploying models on ordinary computers, AI cameras, IoT devices, or other edge devices.

Practical Tips and Conclusion

The workshop concluded with practical technical tips and application experiences shared by young and experienced experts, aimed at helping participants optimize their solutions rapidly. The online workshop successfully equipped the "AI thực chiến" competition participants with a comprehensive knowledge base, covering technological direction, entrepreneurial advice, and in-depth technical tips. The organizers expressed hope that this valuable information would motivate participants, provide them with the necessary knowledge, and offer new perspectives to develop groundbreaking and highly applicable solutions in the competition.