Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore the critical intersection of trustworthy and sustainable artificial intelligence through this comprehensive conference keynote that examines how next-generation computing architectures can address AI's most pressing challenges. Delve into the fundamental limitations of current AI systems, including safety, reliability, and security risks, while understanding the mathematical foundations behind trustworthy AI concepts such as generalization, expressivity, and explainability. Examine AI's growing sustainability crisis, including energy consumption approaching global production limits and the hidden costs of training and large-scale inference operations. Investigate the theoretical limits of digital computing through computability theory and discover why certain inverse problems remain fundamentally non-computable on traditional hardware. Learn about emerging computing paradigms including analog, neuromorphic, and brain-inspired chips that can overcome foundational barriers, with real-world examples like SpiNNaker hardware and its energy advantages. Understand Spiking Neural Networks from mathematical, biological, and efficiency perspectives, and explore how time as a computational dimension opens new AI capabilities. Discover insights from Bavaria's Next Generation AI Computing initiative and the importance of co-evolving hardware and software design for applications in robotics, communication, medicine, and mission-critical systems. Gain a comprehensive vision for developing AI systems that are simultaneously powerful, trustworthy, and energy-efficient through this expert analysis from a leading researcher in mathematical foundations of artificial intelligence and neuromorphic computing.
