AI models are becoming cheaper to run, but rising demand for data centres, computing and electricity is pushing up the infrastructure cost of AI, including in India. According to Business Standard, the falling cost of running individual AI tasks does not necessarily mean lower overall infrastructure demand. In fact, lower costs can encourage businesses to use AI more frequently and across a wider range of applications. Gartner expects worldwide spending on AI-optimised infrastructure as a service (IaaS) to reach $42.3 billion in 2026, up 96.4 per cent from 2025, with total IaaS spending worldwide growing 29.3 per cent in 2026. The efficiency paradox means that more efficient models can reduce computing required for individual tasks, but total demand can continue to rise as AI becomes more affordable for broader business applications.
Krishna Jonnalagadda, global chief technology officer of GE Vernova, believes power is the main constraint in building artificial intelligence data centres. According to reports from Business Standard, Jonnalagadda emphasizes that while data and chips are important for AI, power remains the most critical bottleneck. The company is experiencing significant demand, with orders for gas turbines extending to 2031 due to growing data centre power requirements. GE Vernova is expanding its capacity simultaneously, with output expected to be four times higher than 2019 levels.
The Ministry of Power has projected an additional 26.3 GW of electricity demand from data centres by 2031-32, according to a recent clarification reported by ETEnergyWorld. This represents a significant increase from the earlier estimate of 13.56 GW by 2031-32 cited in March. Of this projected requirement, applications for about 17 GW have already been submitted to state transmission utilities, while another 9.3 GW has yet to enter the formal grid-connection process. The government has said transmission infrastructure is being developed in phases to match growing electricity demand, with 154 inter-state transmission projects under construction involving 39,792 circuit kilometres of transmission lines.
Jonnalagadda identifies AI, robotics and automation as his top three priorities since joining GE Vernova last year. As reported by Business Standard, the company is focusing on energy for AI applications through advanced technologies like solid state transformers and AI for energy operations. The transformative projects include direct air capture (DAC) technology for carbon reduction, solid-state transformers for data centres, and fuel cells that convert fuels into electrons through chemical processes. These initiatives are expected to deliver significant returns on investment, with robotics showing potential for significant ROI by 2028.
AI is delivering a return on investment this year and is expected to grow exponentially. The technology is transforming GE Vernova's operations through three key areas: product development, quality inspection, and grid efficiency. Product development processes that previously took 18 months can now be completed in three months using AI techniques like deep learning surrogates. Quality inspection is being automated through AI and robotics, with robots inspecting blades from all directions to identify defects without human judgment. Grid efficiency improvements are being achieved through software solutions that optimize transmission line cooling and extend to maintaining grid efficiency during peak usage periods.
