Overview
Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore a 23-minute technical talk where Oracle's OCI Chief Technical Architect Pradeep Vincent and Chief AI Scientist Dan Roth examine the fundamental principles of Generative AI engineering and dispel common misconceptions. Delve into the evolution of AI over 25 years, understand how AI models actually learn through sophisticated memorization versus reasoning, and grasp key concepts in spatial reasoning and decision-making processes. Learn about practical approaches to solving real-world problems using systems of models, and discover crucial innovations in AI infrastructure, including challenges in scaling GPU superclusters and power management solutions. Get insights into the projected state of Generative AI in three years and understand how OCI's Generative AI infrastructure is being scaled for enterprise applications. The discussion covers technical aspects from basic AI concepts to advanced infrastructure considerations, making it valuable for both AI practitioners and business decision-makers interested in practical Generative AI implementation.
Syllabus
- Intro to AI and AI Infrastructure Trends
- How Has AI Changed in 25 Years
- Misconceptions about AI
- How AI Models Learn
- Sophisticated Memorization vs Reasoning
- Decision Making: Spatial Reasoning
- Decision Making: Conjunctive Fallacy
- Solving Real-World Problems
- System Of Models
- Gen AI In Three Years
- Innovations in AI Infrastrucure
- Challenges Scaling GPU Superclusters
- Power Management for GPU Superclusters
- Near Future of AI Infrastructure
- Scaling OCI Gen AI for Business
Taught by
Oracle