Next in Tech

This episode of our AI CEO series looks at the potential for AI capabilities to address one of AI's largest challenges – energy availability. Dr. Varun Sivaram, founder and CEO of Emerald AI, joins host Eric Hanselman to explore how greater flexibility is AI workloads can unlock unused capacity. Traditionally, data center construction has focused on delivering peak capacity, even at those times when it may not be needed. That doesn't mesh well with a power grid that is already under stress. AI intelligence can shift workloads dynamically to reduce data center demands when the power grid needs it and to leverage excess grid capacity when it's available.
Flexing AI workloads to respond to grid conditions isn't simple, but it can be more attractive than alternatives. Adding generation capacity is a long-term and expensive process and community concerns about rate increases are creating headwinds for new data center builds. Adding Battery Energy Storage Systems (BESS) are costly and have their own capacity and durability limitations. Flexing workload demand could address grid integration problems on a much shorter timeline.
More S&P Gobal content:
2026 Trends in Data Center Services & Infrastructure
Geopolitics of data centers: An AI showdown that will reshape the world
AI's global resource race: Challenges and opportunities
Hyperscaler earnings quarterly: What price inference?
For S&P Global subscribers:
Data Center Services & Infrastructure Market Monitor & Forecast: US ISO Region
BYOP emerges as a strategic decision for data center operators in response to grid challenges
US Independent System Operators 101
Service providers race to meet surging enterprise demand for AI infrastructure
Survey Data Hub – VoCUL: Connected Customer, AI & Data Centers 2026
Host/Author: Eric Hanselman
Guest: Varun Sivaram
Producer/Editor: Dylan Scheible
Published With Assistance From: Feranmi Adeoshun and Sophie Carr

Market dynamics in data center buildouts and dealmaking are complex, but there are some common threads that are spanning the globe. Analysts Soon Chen Kang and Stefanie Williams join host Eric Hanselman to explore their research in the Asia Pacific market. AI demand is a driving force and concerns about power availability can be a headwind, but there are regional nuances that make the markets that comprise the APAC region unique. One treats the market as a cohesive whole at one's peril!  Geopolitical concerns have a significant impact on both demand forces and capacity uptake. Access to GPU's to feed AI applications takes on new dimensions in a market where technology embargoes are moving workloads. Neoclouds are a new market provider and driving new builds, as well.
The boom days of speculative building have tapered off, in many cases, due to the availability of power, but that doesn't mean activity is slowing. Activity is shifting to find available capacity, but a common challenge in some of these early markets is the lack of transparency on both power deals and the build and financing processes. Check out the research to get more details.
More S&P Global Content:
Compute sovereignty: The strategic importance of digital infrastructure
AI in action: unleashing agentic potential
Hyperscaler earnings quarterly: What price inference?
Will datacenter growth in India propel it to global hub status?

For S&P Global subscribers:
Beyond disaster recovery: Greater Osaka's rise as a stand-alone data center hub
APAC data center growth on course in 2026, driven by hyperscaler, regional neocloud demand
Rising community opposition impacts site selection, cancellations and delays increase 
Singapore kicks off tender process, promising at least 200 MW for data centers
2026 Trends in Data Center Services & Infrastructure
Host/Author: Eric Hanselman
 
Guests: Soon Chen Kang, Stefanie Williams
Producer/Editor: Dylan Scheible 
Published With Assistance From: Feranmi Adeoshun, Sophie Carr, Kyra Smith,

We're in the early days of cost impacts for AI applications. While there are some cautionary tales, current spending seems to be a small fraction what's to come. Analysts Jean Atelsek and Melanie Posey return to the podcast to talk about what they heard at the FinOps X conference with host Eric Hanselman. The need for cost management in AI is seen as so great that the FinOps Foundation, a project of the Linux Foundation, is talking about morphing its conference into Tokenomicon and pivot into token economics. The portmanteau of tokenomics is sweeping across cloud and AI services providers, as well as IT vendors, as enterprises wrestle with dueling forces of AI acceleration and management constraints for access and cost.
Unlike FinOps for cloud operations, the costs and metrics for AI are fairly opaque. Some enterprises are trying to manage costs by limiting access, but that risks stifling the innovation and democratization that is supposed to come with AI transformation. Request routing is promising, but it requires understanding the nature of the request and the suitability of available infrastructure to fulfill it, something that is not well understood by many. Most are just getting comfortable with managing model lifecycles and the step up to cost management can be a large one.
More S&P Global Content:
Compute sovereignty: The strategic importance of digital infrastructure
Next in Tech | Ep. 222: FinOps – Managing Cloud and AI Costs
AI in action: unleashing agentic potential
Hyperscaler earnings quarterly: What price inference?
For S&P Global subscribers:
FinOps in the age of agentic AI
FinOps Foundation expands FinOps discipline beyond cloud to technology value management
Service providers race to meet surging enterprise demand for AI infrastructure
FinOps Market Monitor & Forecast
Host/Author: Eric Hanselman 
Guests: Jean Atelsek, Melanie Posey
Producer/Editor: Dylan Scheible
Published With Assistance From: Feranmi Adeoshun, Sophie Carr, Kyra Smith,

There's great momentum in moving to greater levels of agentic automation, but there are critical areas where deeper consideration is required in how it's applied. In capital markets, trust is a foundational element on which transactions are built and Krisha Vinjamuri, Head of Technology, Enterprise Solutions at S&P Global Market Intelligence, joins host Eric Hanselman to talk about how this can be achieved and the important aspects of successful implementations.
One of the useful things in capital markets, is that there are open standards on which to base data ontologies. It's not exciting, but it's the basis of a semantic foundation that can not only ensure that there is depth in data definitions, but can also reduce errors generated by agents. The larger question that looms beyond the construction of foundational architecture, is how the operational envelope that bounds agentic action will be established. This has to be built from policy definitions that take those actions into account. There is great promise and much work that needs to be done.
More S&P Global Content:
AI-enable Corporate Actions
AI-driven Portfolio Monitoring with iLEVEL
AI-automated SSI
Equity Bookbuilding AI Assist
Market Intelligence AI Hub
For S&P Global subscribers:
FinOps in the age of agentic AI
AI Infrastructure Market Monitor & Forecast
Service providers race to meet surging enterprise demand for AI infrastructure
In 2026, the telecom network becomes code
Credits:
Host/Author: Eric Hanselman
Guest: Krishna Vinjamuri
Producer/Editor: Feranmi Adeoshun
Published With Assistance From: Sophie Carr, Kyra Smith, Dylan Scheible

Networking can be an invisible part of IT infrastructure, but AI is creating demands that make it a critical part of keeping AI application fed with data. Mike Fratto returns to the podcast to discuss both the long haul and local requirements for AI networking with host Eric Hanselman. It's always been important to link chunks of infrastructure efficiently, but AI's voracious need for data has dramatically increased the scope and scale of the need. The risk that any gap in performance or capacity presents is that precious GPU resources will be idled, an increasingly expensive proposition.
The realities of AI application architectures is that infrastructure is ever more hybrid, requiring access to repositories of data both on-premises and in various clouds and models scattered across various providers. The need for dynamic connectivity is driven by the rapid evolution of preferences for new models and the diversifying needs of agents to reach new data sources. It's not only forcing network expansion, but it's also driving M&A activity as network providers look to enhance automation in response to customer demands.
More S&P Global Content:
Compute sovereignty: The strategic importance of digital infrastructure
AI won't solve its own energy problem – and that might be fine
AI in action: unleashing agentic potential
AI infrastructure results in 2025 top expectations, forecast upgraded
For S&P Global subscribers:
MWC 2026: Agentic AI as the next operating model for networks and network operations
AI Infrastructure Market Monitor & Forecast
Service providers race to meet surging enterprise demand for AI infrastructure
In 2026, the telecom network becomes code
Credits:
Host/Author: Eric Hanselman
Guest: Mike Fratto
Producer/Editor: Feranmi Adeoshun
Published With Assistance From: Sophie Carr, Kyra Smith, Dylan Scheible