1991: Linus Torvalds announces Linux

On 25 August 1991, a 21-year-old Finnish computer-science student named Linus Torvalds posted a short message to the comp.os.minix newsgroup. He explained that he was writing a free operating system “just a hobby,” something that would not be “big and professional like gnu.” That unassuming note is now recognised as the public birth of Linux.

Torvalds had grown frustrated with the limitations of Minix on his new Intel 386 PC and set out to build a lean Unix-like kernel of his own. He released the source code under a licence that encouraged sharing and improvement. Within months a small circle of volunteers joined him; within a few years Linux was running web servers, databases and scientific workloads around the world.

The link to artificial intelligence is infrastructural yet decisive. Training today’s large models demands thousands of tightly coordinated processors, enormous memory and reliable networking. The great majority of those machines—whether in university labs, cloud data centres or corporate GPU clusters—run Linux. Container systems such as Docker, orchestration tools such as Kubernetes, and the dominant deep-learning frameworks all grew up in the Linux ecosystem.

Equally important is the cultural precedent. Linux proved that a complex, production-grade system could be built by a globally distributed community working in the open. That model of collaborative development later shaped the release of model weights, training code and datasets that accelerate modern AI research.

What began as a student’s late-summer experiment thus became the quiet operating layer beneath the AI boom. Every time a new language model is trained or a robot’s control stack is deployed, the odds are high that Linux—announced on this day in 1991—is somewhere in the stack.