As the founder said: “AI is a tool and will not replace programmers. Just as compilers eliminated handwritten assembly, they did not eliminate programmers; instead, they allowed more software to be written.”
In today’s world where AI tools are becoming increasingly prevalent, operations roles are not being replaced but are instead experiencing a value upgrade.
Just a few days ago, at the Linux Foundation Open Source Summit held in South Korea, Linus had an interview where he emphasized: “The most important part of software is not innovation, but long-term maintenance.” This precisely highlights the core value of Linux operations. AI can indeed efficiently handle repetitive tasks: automatically tracking upstream versions, generating security patches, and resolving dependencies, significantly improving package management efficiency.
However (nothing is absolute, right?), when systems encounter sudden failures or face complex scenarios of cross-branch synchronization, the limitations of AI become glaringly apparent.
In the same interview, Linus complained that “AI-generated garbage bug reports have become a new DoS attack in the open-source community,” precisely because these tools lack a deep understanding of the underlying logic of systems, which is the core competitive advantage of operations engineers.
The value of operations engineers is more reflected in their ability to manage complexity. Linus bluntly stated that AI will eliminate “those who can only write a bit of code”; what is needed in the future are experts who can grasp the overall system. The success of the Linux kernel stems from its “boringly” stable characteristics, and this stability relies on operations engineers making precise trade-offs between compatibility and functionality during version iterations, making decisions with the aid of automation tools. Just as Nvidia has become a core contributor to the kernel due to AI demand, operations engineers must also leverage AI tools to focus their efforts on architecture design, risk prediction, and other critical areas.
The relationship between AI and operations engineers is no longer one of replacement but of collaboration. Linus mentioned that he “hasn’t really written code in 20 years,” yet he remains the captain of the kernel, indicating that the core of operations is system control rather than repetitive tasks. AI is responsible for basic tasks such as patch generation and environment configuration, while engineers focus on building a trustworthy system supply chain and optimizing large-scale cluster performance. This division of labor is akin to the collaboration between compilers and programmers, where tools free up hands, and talent creates value.
Linus’s calmness cools the industry: “Current AI is still far from good enough.” For Linux operations, AI is an enhancement tool rather than a replacement. Those who can master AI tools, delve into system maintenance, and possess a global perspective will become essential in the industry. Just as compilers have spawned more software developers, AI will continuously expand the boundaries of Linux operations, making truly valuable talent increasingly irreplaceable—this is the insight Linus offers to all operations personnel.
So, do not hesitate about what to learn; you must learn what you need to learn. This is very similar to a topic from years ago: what language should one learn? Many people wasted time on the choice. In fact, there aren’t that many choices; just do it!