About the webinar:
Why do today’s AI systems still struggle to act autonomously in the real world? While models like ChatGPT and advanced vision systems perform well at perception and pattern recognition, they lack the ability to predict, imagine, and plan, which are key ingredients of true intelligence.
In this webinar, we explore how world models are transforming AI from reactive systems into predictive, goal-driven agents. The webinar will introduce the core concepts behind world models, explain how approaches like Joint Embedding Predictive Architectures (JEPA) shift learning from pixel-level prediction to meaningful representation learning, and highlight why this matters for scalability and real-world performance.
Finally, we will dive into V‑JEPA 2, a state-of-the-art self-supervised model that learns from raw observation and enables planning capabilities without explicit supervision. Through this lens, we will discuss how AI can move closer to human-like understanding and autonomy and what this means for the future of robotics and intelligent systems.
Who is the webinar for?
-
Researchers and practitioners in AI, machine learning, and robotics
-
Engineers interested in autonomous systems, simulation, and decision-making
-
Professionals exploring the next generation of AI beyond large language models
-
Students and academics looking to understand world models and predictive intelligence
-
Anyone curious about how AI can move from perception to planning and real-world action
Key takeaways for participants:
-
Why current AI systems struggle with true autonomy and the limitations of perception-only intelligence
-
How world models enable prediction, reasoning, and planning in intelligent systems
-
The shift from pixel-level prediction to representation learning using approaches like JEPA
-
How learning from large-scale video data supports understanding of real-world dynamics
-
Challenges and future directions toward building scalable, autonomous AI systems
Speaker bio:
Sarder Fakhrul Abedin received his B.S. degree in Computer Science from Kristianstad University, Sweden, in 2013, and his Ph.D. degree in Computer Engineering from Kyung Hee University, South Korea, in 2020, supported by the President Scholarship and the Brain Korea (BK) 21+ program.
Currently, he is a Senior Researcher with the Connected Intelligence Unit at RISE Research Institutes of Sweden, Kista. In addition, he serves as an Adjunct Lecturer in the Department of Computer and Electrical Engineering at Mid Sweden University, Sundsvall, where he previously held a position as Senior Lecturer.
Throughout his career, he has led multiple AI and machine learning–driven research projects, holds several patents, and has authored more than 50 peer-reviewed publications. As an IEEE Senior Member, he actively contributes to standardization efforts and is engaged with the IEEE Communications Society’s Technical Community on Intelligent Informatics and Computer Communications. He also serves as a Section Board Member for MDPI Sensors.
His research interests include autonomous systems, network intelligence, edge computing, and explainable AI.