Abstract

The AI revolution has unfolded at an unprecedented pace, defying even the most optimistic predictions. From AlexNet’s breakthrough in 2012 to the Transformer architecture’s debut in 2017, each milestone has accelerated the field’s advancement. Today’s landscape features sophisticated large language models like GPT-4, Claude, and Gemini, alongside powerful multimodal AI systems that seamlessly integrate text, vision, and other modalities.

This seminar traces AI’s evolution, with particular emphasis on the transformative developments of the past decade. We’ll explore how foundational deep learning breakthroughs have disrupted industries across the technological spectrum. Special attention will be given to the emergence of large language models and their enabling of versatile multimodal AI agents, as well as parallel advances in robotics. We’ll examine current research trends and analyze how these technologies are reshaping products and services across sectors.

The discussion will conclude by exploring the intersection of AI and cryptography, specifically focusing on business opportunities for homomorphic encryption in the AI era. This presents an exciting opportunity for collaboration with the Lyon Research Center of CryptoLab researchers to identify promising applications and use cases.

Supplementary materials will be provided as appendices, covering additional topics including AI hardware developments, AI applications in biotechnology, and detailed analysis of LLM and generative AI technology markets. While these topics won’t be covered during the main presentation, they offer valuable context for further exploration.