In a thought-provoking discussion at the Milken Global Conference, five key figures in the AI industry shed light on the challenges and opportunities shaping the future of artificial intelligence. From chip shortages to the exploration of orbital data centers, these experts delved into the very foundations of the AI economy, offering unique insights and raising intriguing questions.
Bottlenecks and Beyond
Christophe Fouquet, CEO of ASML, highlighted the urgent need for increased chip manufacturing, predicting a supply-limited market for the next few years. This constraint, coupled with Francis deSouza's emphasis on Google Cloud's rapid growth and massive backlog, underscores the urgency of addressing the physical limits of AI.
Qasar Younis, CEO of Applied Intuition, introduced a different bottleneck: the data required to train AI models. His perspective emphasizes the importance of real-world data collection, a challenge that extends beyond silicon.
Energy: The Next Frontier
The energy problem looms large, with deSouza revealing Google's exploration of data centers in space as a response to energy constraints. This innovative approach highlights the industry's commitment to efficiency and integration, as seen in Google's co-engineered AI stack. Fouquet's emphasis on the cost of energy further underscores the need for sustainable solutions.
Redefining Intelligence
Eve Bodnia, a quantum physicist-turned-entrepreneur, challenges the status quo with her startup, Logical Intelligence. Her approach, based on energy-based models, offers a unique perspective on AI, arguing that it aligns more closely with human brain function. With a model running at thousands of times the speed of leading LLMs, Bodnia's work raises questions about the sufficiency of scale in AI.
Agents and Control
Dimitry Shevelenko, chief business officer of Perplexity, discussed the evolution of their search product into a 'digital worker,' raising important questions about control and security. The concept of 'granularity' in permissions and the need for approval before actions are taken by agents are crucial considerations in ensuring the responsible development and deployment of AI.
Sovereignty and Physical AI
Younis' observation about the geopolitical implications of physical AI is particularly intriguing. The entanglement of physical AI with national sovereignty raises questions about safety, data collection, and control. Fewer nations can field robotaxis than possess nuclear weapons, highlighting the potential impact of AI on global power dynamics.
The Future of Work
The panel's discussion on the impact of AI on the next generation's critical thinking skills offers a hopeful perspective. DeSouza's emphasis on the potential to address complex problems is balanced by Shevelenko's focus on accessibility and Younis' distinction between knowledge work and physical labor.
In conclusion, this insightful discussion highlights the multifaceted nature of the AI economy, from technical challenges to geopolitical considerations. As the industry navigates these bottlenecks and explores new frontiers, the future of AI promises to be both exciting and transformative.
