Uber CEO Dara Khosrowshahi: AI Transforms Engineers into ‘Superhumans’ – But in 5 Years, NRIs May Face Tough Choices as ROI Shifts to AI Agents and GPUs
For Non-Resident Indians (NRIs) in the software engineering space—especially those on H-1B visas in the US or building careers in tech hubs like Silicon Valley, Seattle, or New York—Uber CEO Dara Khosrowshahi’s recent revelations hit close to home. In a February 2026 appearance on The Diary of a CEO podcast with Steven Bartlett (aired around February 23, 2026), Khosrowshahi shared eye-opening insights into how artificial intelligence is reshaping engineering productivity at one of the world’s most code-dependent companies.
Khosrowshahi revealed that 90% of Uber’s software engineers now integrate AI tools into their daily workflows. Among them, the top 30%—the “power users”—are experiencing unprecedented leaps in output. These elite adopters are submitting significantly more “diffs” (code changes or pull requests) to Uber’s vast codebase, a key indicator of velocity in a platform that powers global ride-hailing, delivery, pricing algorithms, fraud detection, and more.
“About 90% of our coders are using AI,” Khosrowshahi explained. “Now, that’s easy to say, but there are probably 30% of them that are power users. They are showing a clear differentiation in the number of diffs… how much a diff is a code release that’s different from the last code release. So one of the measurements of productivity is just how many diffs are you putting to the code base.”
This isn’t just incremental improvement—it’s a paradigm shift. Power users orchestrate AI agents for repetitive tasks, debugging, architecture exploration, and even real-time system monitoring, freeing humans for high-level strategy, innovation, and oversight. Khosrowshahi highlighted how AI agents now proactively diagnose issues across systems, slashing the need for large on-call rotations that once consumed hours of manual troubleshooting. “The human can look over the shoulder of the AI agent,” he noted, describing a collaborative future where engineers become overseers of intelligent systems.
For NRIs, who form a significant portion of the US tech workforce (with Indians holding a large share of H-1B visas in software roles), this current phase is largely positive. Khosrowshahi emphasized that AI isn’t prompting headcount reductions at Uber—in fact, the opposite. “I just think they become superhumans,” he said in earlier 2025 comments echoed in the podcast. “So we are actually hiring more engineers because every engineer got more valuable to me.” The company is leveraging these gains to accelerate innovation, from recommendation engines to autonomous vehicle tech, rather than shrinking teams.
Yet the forward-looking prediction carries stark implications for NRI software professionals. Khosrowshahi forecasted a tipping point in roughly five years (around 2031), when the return on investment (ROI) for hiring an additional human engineer will be outpaced by scaling AI agents and investing in GPU compute power—particularly from NVIDIA.
“At that point, instead of adding an engineer, I should add agents and buy some more GPUs from Nvidia,” he stated bluntly. As agentic AI evolves—handling longer unsupervised tasks, multi-step planning, and autonomous iteration—the marginal cost of deploying another “agent” plummets compared to salaries, benefits, relocation, and visa sponsorship for humans. For companies like Uber, this economic logic could shift hiring priorities dramatically.
This resonates deeply with NRIs. Many Indian-origin engineers in the US have built careers on specialized skills, H-1B sponsorship, and pathways to green cards. But as AI augments (and potentially substitutes) routine and mid-level coding, the demand for “superhuman” talent—those mastering AI orchestration, system design, and emerging domains—will concentrate at the top. Entry- and mid-level roles, often filled by fresh graduates or offshore transitions, face greater automation risks, mirroring trends in India’s IT sector where AI already handles 20-40% of development tasks in many firms.
Broader context adds urgency. Uber’s push into robotaxis and autonomy already threatens millions of driver jobs globally, including many held by NRIs or recent immigrants. Meanwhile, in the US tech ecosystem, AI-driven productivity could chill sponsorships for routine engineering roles, intensifying competition for premium positions. Khosrowshahi himself noted AI’s potential to disrupt 70-80% of knowledge work over the next decade, urging preparation for massive societal shifts.
For NRIs, the message is dual-edged: seize the “superhuman” opportunity now. Upskill aggressively in AI agents, prompt engineering, ML ops, autonomous systems, and domain-specific applications. Those who become power users—leveraging tools to multiply output—will remain indispensable. But the five-year horizon warns of a pivot: companies may favor scalable compute over scalable headcount, potentially reducing demand for traditional software engineering immigration pathways.
Uber continues to hire aggressively in 2026, with Khosrowshahi betting on human-AI synergy for the near term. Innovations like “Dara AI”—an internal chatbot clone engineers use to rehearse presentations—show how deeply AI embeds even in leadership prep. Yet his candid outlook serves as a wake-up call for NRIs: adapt swiftly, master AI as a multiplier, or risk the economics tilting toward silicon over people.
As frontier AI accelerates, the NRI tech community—long a pillar of global innovation—must navigate this transition thoughtfully. The superhuman era is here, but the agent-and-GPU era looms. For engineers abroad building futures for families back home, the next five years will test resilience, adaptability, and strategic upskilling like never before.
