Linus Torvalds: AI Is Both Pain and Power in Linux Kernel, but Not a Cure-All
At Open Source Summit in Mumbai, Linus Torvalds discussed the dual role of AI in Linux kernel development. AI helps: quickly write code, find errors, generate tests. But Torvalds is critical — AI doesn't understand Linux context, may suggest suboptimal solutions. Plus, Linux removes support for old technologies — 'museum' architectures that take up code space but are rarely used.
AI-processed from ZDNet AI; edited by Hamidun News
Linus Torvalds, creator of Linux, gave an interview at Open Source Summit. The main theme: AI in the Linux kernel is both a pain and a power.
The pain of AI
AI in kernel development can be dangerous:
- Context: AI doesn't understand code history, why a particular module was written
- Optimization: may suggest code that runs fast on new processors but slowly on old ones
- Architectural decisions: AI writes according to patterns from training data, but the kernel requires conscious architectural choices
Torvalds warns: AI is not a replacement for understanding.
The power of AI
But there are benefits:
- Fast boilerplate generation (driver templates, config parsers)
- Error detection through pattern matching (memory leaks, race conditions)
- Test generation
- Automation of tedious work
Abandoning old technologies
Torvalds also confirmed: Linux is stopping support for "museum" architectures. Why keep code in the kernel for processors that nobody uses? It slows down development, complicates testing, wastes memory.
Examples of what's being removed: support for 32-bit ARM (even old screens use 64-bit), specific platforms that are dead.
What it means
Torvalds takes a pragmatic view: AI is a tool, not a panacea. Linux remains open, but requires understanding from developers. Dropping old code is an acknowledgment: evolution requires sacrifices. An optimized kernel, even if it stops supporting 10-year-old chips, is a savings for 99% of users.
Key Questions Answered
What are the key details on The pain of AI?
AI in kernel development can be dangerous: Context: AI doesn't understand code history, why a particular module was written Optimization: may suggest code that runs fast on new processors but slowly on old ones Architectural decisions: AI writes according to patterns from training data, but the kernel requires conscious architectural choices Torvalds warns: AI is not a replacement for understanding.
What are the key details on The power of AI?
But there are benefits: Fast boilerplate generation (driver templates, config parsers) Error detection through pattern matching (memory leaks, race conditions) Test generation Automation of tedious work
What are the key details on Abandoning old technologies?
Torvalds also confirmed: Linux is stopping support for "museum" architectures. Why keep code in the kernel for processors that nobody uses? It slows down development, complicates testing, wastes memory. Examples of what's being removed: support for 32-bit ARM (even old screens use 64-bit), specific platforms that are dead.
Build & Ship Production AI Agents Without Boilerplate Code
Start your free 72-hour trial at Hamidun Academy (academy.hamidun.com) — interactive lessons on Claude Code & Cursor, live Tech Radar, and production agent templates.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.