Twenty years ago, Raghu Nambiar and Debo Dutta were colleagues at Cisco, filing patents together and tinkering with GPUs long before the world had heard of generative AI. Today, Nambiar leads the data center engineering team at chipmaker AMD, and Dutta serves as chief AI officer at Nutanix. Their reunion at the AMD Advancing AI event in San Francisco revealed a shared conviction: agentic AI will outstrip every transformational technology that came before it.
"If you look at last 50 years, we have seen many transformational technologies like the internet, big data," Dutta said, standing next to Nambair during a video interview with The Forecast.
"But agentic AI is going to be way more transformational than anything that we have seen in the past. Every tool that people are re-imagining in this era is basically running AI agents underneath.”
The market is racing to catch up with that ambition. Gartner predicts 40% of enterprise applications will feature task-specific AI agents by the end of 2026, up from less than 5% in 2025. The agentic AI market is projected to grow from $10.41 billion in 2025 to $52.6 billion by 2030, according to Kearney. Yet Forrester reports that while three-quarters of enterprise leaders say they are adopting agentic AI, only a small minority have it running in meaningful production.
Three Forces Converge
Nambari traced the breakthrough to three converging forces: algorithms that matured over 50 years, explosive compute power from CPUs and GPUs, and the data infrastructure built during the big data era.
"Combining that with the hardware technology, CPU and GPUs and the algorithms that matured over the last 50 years is enabling agentic AI," said Nambiar, AMD corporate vice president of engineering,leading compute, enterprise and sovereign AI, and high performance computing (HPC).
Dutta said, who incubated Nutanix Enterprise AI and now drives the company's AI strategy, sees a strong connection to the past. The science fiction vision of mimicking human intelligence is decades old, he said. What changed is pairing that intelligence with the automation.
"If you combine that with whatever jagged intelligence we have, it can become really powerful," Dutta said.
From Pilots to Productivity
At AMD, a Fortune 500 company, the theory is already practice, explained Nambiar. He oversees a global engineering team supporting more than 17,000 developers, and agentic AI is embedded in development, testing and quality assurance. The target is total adoption.
"Our target is to have 100% of the developers using agentic AI tools," Nambiar said. "And we expect a significant improvement in productivity in coming months. Not coming years, coming months."
Nutanix is pursuing the same productivity push internally, Dutta said, applying agents across prototyping, design and research.
“Even within Nutanix, our target is to get more productive, drive productivity through the use of AI and agents,” Dutta said.
The Human Premium
The tools change; the need for sharp people does not. Dutta said the bar has only risen.
"You needed smart people back then. Now you need smarter people," he said. "In order to operate these powerful agents, you need to build really strong teams."
Where standard intelligence gets commoditized, humans bring intuition, empathy and the ability to connect dots, Dutta said. He sees AI as a lever that lets people think bigger.
"I actually believe that AI will enable us human beings to actually think 10X bigger and then solve extremely hard problems where we are still stuck," he said.
Nambiar agreed, noting that the next generation has access to tools and data that earlier engineers spent decades waiting for.
"The pace of innovation in next few years is going to be something like what we have achieved in a few decades," he said. "The pace of innovation is going to be tremendous."
An Agent for Everyone
The conversation closed on a democratic note. It took years for mobile phones to reach seven billion people, Nambiar said. Personal AI agents will spread faster, he predicted, reaching healthcare, finance and vacation planning.
"Pretty soon everybody in the world is going to have agents," he said.
Dutta wants more than one.
"I think one agent is not good enough for me. I need a few," he said.
The real challenge, he added, is ensuring everybody benefits.
"For us technology leaders, we have a big role to play in making that happen."
Related:
Jason Lopez is executive producer of Tech Barometer, the podcast outlet for The Forecast. He’s the founder of Connected Social Media. Previously, he was executive producer at PodTech and a reporter at NPR.
Ken Kaplan contributed to this story. He is Editor in Chief for The Forecast by Nutanix. Find him on X @kenekaplan and LinkedIn.
© 2026 Nutanix, Inc. All rights reserved. For additional information and important legal disclaimers, please go here.
Related Articles
