Dr. Lav Varshney is currently the AI Institute Director at Stony Brook University and is the Chief Scientist for Ensaras, Inc. He has held teaching positions at UIUC and Northwestern. Additionally, he was a principal research scientist at Salesforce Research AI and a former White House Fellow on the National Security Council. 

 

I spoke with Varshney to learn more about his perspective on these issues, as he has experience in the academic, business, and political sides of this issue. 

 

Varshney emphasized the importance of transparency with data centers. Having a clear understanding through transparent usage decomposition in multi-tenant data centers is important for developing effective sustainability solutions. Varshney recommends looking at the full lifecycle of AI, including the semiconductor supply chain and the physical infrastructure of data centers. He believes that transparency will help people better understand what is occurring and enable informed policymaking. 

 

In a previous interview, Varshney discussed the Jevons paradox in the context of developing efficient AI technologies. The Jevons paradox is the idea that when technology increases efficiency in using a resource, the resulting lower cost causes increased demand, leading to an ultimate increase in consumption. Varshney explained that to combat this, policy solutions might focus on decoupling efficiency from cost reductions. The examples he provides are setting absolute limits on usage or progressive taxes.