AI boosts developer work, not replacing jobs: Alphabet CEO

Google CEO Sundar Pichai told American computer scientist, Lex Fridman, that AI is not replacing developers but helping them enhance their work.

He said tools like Google’s Gemini reduce repetitive coding tasks, allowing engineers to focus more on creative and problem-solving activities. About 30% of the code written at Google is now supported by AI.

Pichai added that AI has boosted engineering velocity by around 10%, improving how fast and efficiently new features and updates are delivered.

Google is still hiring engineers, as AI creates new opportunities and allows more people to engage in meaningful and creative work.

🔗 Source: The Economic Times

🧠 Food for thought

1️⃣ Engineering velocity emerges as the true ROI metric for AI coding tools

Pichai’s focus on “engineering velocity” rather than cost-cutting aligns with broader industry trends measuring AI’s impact on development speed.

Google’s internal measurements show a 10% improvement in engineering velocity from AI adoption, demonstrating that organizations are quantifying productivity gains beyond simple code generation.

This mirrors Block’s approach, where they’re tracking “DEVIQ” metrics (developer experience, velocity, impact, and quality) and aim to reclaim over 500,000 developer hours annually through AI and workflow improvements 1.

The industry is increasingly viewing AI as a productivity multiplier rather than a replacement technology, with companies like GlobalLogic launching specialized platforms like VelocityAI specifically to enhance development speed 2.

GitHub’s surveys show 92% of developers now use AI coding tools, indicating mainstream adoption focused on acceleration rather than replacement 3.

This productivity-centered approach explains why Pichai confirms Google will continue hiring engineers despite AI advancements. The goal is expanding capabilities, not reducing headcount.

2️⃣ The programmer’s role shifts from code writer to code director

Pichai’s revelation that 30% of Google’s code is AI-assisted reflects a fundamental transformation in how software is created across the industry.

Engineers are increasingly becoming “editors” rather than authors of code, focusing on system architecture and design decisions while delegating implementation details to AI 4.