Advancing AGI: From Training and Fine-Tuning to Interactive Agent Development

ACM Distinguished Speaker Series
Title: Advancing AGI: From Training and Fine-Tuning to Interactive Agent Development
Speaker: Mehdi Bahrami – Santa Clara, CA, United States
Date/Time: Friday, April 17 | 11:30 AM – 1:00 PM

As we continue the ACM Distinguished Speaker Series, Month 3 explores one of the most ambitious frontiers in computing: the advancement of Artificial General Intelligence (AGI) and the evolution from model training to fully interactive, adaptive agents.
This session traces the historical foundations of AGI and examines where the field is heading next. Beyond improving model performance, AGI aims to enable systems that can autonomously interact, reason, and perform unseen tasks in complex, dynamic environments.

Dr. Bahrami will guide us through the critical stages of the AGI journey, including:

• Training and fine-tuning large language and vision models
• Optimization and alignment techniques
• Multimodal integration across text, vision, and beyond
• Designing interactive agents that learn, adapt, and reason across domains
• Practical strategies for safety, reliability, and long-term coherence in autonomous systems
• Societal and technical challenges in pursuing fully autonomous AGI

This talk emphasizes both technical depth and responsible development, aligning with the Responsible Tech Forum’s commitment to advancing innovation while addressing ethical, societal, and professional considerations.

Who Should Attend
AI engineers, data scientists, researchers, product leaders, students, and professionals interested in advanced AI systems, agent design, and the responsible evolution of AGI.

What You’ll Gain
By the end of the session, attendees will:

• Understand the architectural and optimization pathways from foundational models to interactive agents
• Gain insight into cutting-edge techniques shaping AGI development
• Develop a deeper perspective on the societal and safety implications of autonomous AI systems
• Be equipped to think critically about building scalable, reliable, and responsible AI agents

This session continues our structured progression from foundations to applied agent systems, thereby building technical rigor while placing responsibility at the center of next-generation AI development.

About Responsible Tech
Responsible Tech is an ACM-sponsored forum for practitioners examining how AI systems are designed, deployed, and governed in real-world contexts where trust, safety, and accountability matter.

This event is virtual and will be held from 11:30 AM to 1:00 PM US Central Time (CST).