Conference Presentation, Keynote
Yuri Frayman, Co-Founder & CEO, Cast AI: Robots Will Feed the World… If We Build the Rails
- The organization plans to deploy infrastructure on a planetary scale over the next 10 years, including purpose-built GPU-intensive platforms like "Project Stargate," data centers, and infrastructure "rails" deemed essential to feed the world and democratize AI access.
- To power future operations for billions of intelligent devices generating petabytes of data, the organization intends to produce every possible micron of gatherable energy, asserting that nothing will scale without rethinking electron creation.
- The entity will leverage existing technology to impact every currently existing person and industry, aiming to redefine sectors through cloud-based AI training and fine-tuning rather than local processing.
- Operations will require significantly more GPU power to move AI into the physical world, facilitated by a distributed architecture designed to provide affordable, zero-friction, cloud-agnostic inference services that keep models and customer data secure within private control environments.
- The workforce must undergo re-education to reach resource abundance, and the organization will need to produce more AI and ML engineers to maintain necessary human oversight, countering the myth that AI destroys jobs by citing historical trends of prosperity from technological shifts.
- The organization forecasts that while a small number of individuals may build models to decouple regulatory constraints to aid younger companies, the future AI ecosystem will be open, fast, and fair, with the infrastructure rails of the coming decade determining who benefits from this global economic growth.