Microsoft President: AI Will Change, Not Replace, Engineers
Microsoft President Brad Smith stated that AI will not replace engineers but will change the nature of their work. The comments follow the company's 2025 layoffs of over 10,000 employees, which were framed as a move to prepare for the 'AI era'. Smith's perspective suggests AI will serve as a force multiplier, requiring engineering teams to upskill in AI and workflow automation.
- Microsoft's 2025 layoffs, which affected over 15,000 employees in the first seven months of the year, were not a cost-cutting measure due to financial distress but a strategic reallocation of capital. The company is redirecting funds from headcount to an $80 billion investment in AI and data center infrastructure across 40 countries. - Brad Smith has also voiced concerns about US export rules limiting AI component sales to allies, including India, arguing it could push customers toward Chinese AI suppliers and hinder the global diffusion of technology. He has urged the industry to learn from the uneven distribution of benefits during past industrial revolutions to ensure AI technology is accessible to everyone. - The shift in engineering work is already impacting hiring, with a Stanford study indicating a 13% relative decline in employment for early-career engineers (ages 22-25) in roles most exposed to AI. The perspective from a Morgan Stanley analyst suggests that as AI handles more routine coding, the demand for senior developers who act as curators, reviewers, and problem-solvers will grow. - The adoption of AI developer tools is widespread, with the 2025 Stack Overflow Developer Survey revealing that 84% of developers now use or plan to use them. This ecosystem includes not just code assistants like GitHub Copilot but also AI-first code editors like Cursor and full-app builders like Vercel's v0, which generates front-end code from text prompts. - The next frontier is "agentic AI," where tools like Devin AI are being developed to manage entire software projects from end to end, moving beyond simple code completion to task execution. This reflects a shift from AI as a passive assistant to a more autonomous collaborator in the development lifecycle. - In the Bangalore startup ecosystem, several companies are building for this new era of development. Tracxn lists over 16 AI companies in the application development tools space locally, including Appsmith, which provides an open-source low-code platform, and LatentForce.ai, a seed-stage startup building AI agents for software products. - Discussions on developer forums like Hacker News reflect this transition, with a consensus that AI is far from replacing engineers on complex software that requires deep domain knowledge. The sentiment is that AI acts as "autocompletion on steroids," handling a percentage of the implementation phase, while the core engineering work of architecture, problem-solving, and system design remains a human task. [cite: 30
