What are Bottlenecks to Scaling AI?

Widescale AI adoption will require overcoming a handful of hurdles. Owen Hyde, Managing Director and Technology Equity Research Analyst at Jennison Associates, explores challenges such as the significant amount of electricity required to power data centers, the need for synthetic data, and the use of distributed training.

format_quote
As we're thinking about the requirements to get to the next level of model, [energy] is one of the biggest constraints.
As we're thinking about the requirements to get to the next level of model, [energy] is one of the biggest constraints.

Hyperscale

The Artificial Intelligence and Data Center Revolution