The IT industry must continuously innovate to survive. This article analyzes the development of the Linux kernel through the lens of biological evolution, suggesting that selective pressures drive technological change. Three microkernel projects demonstrate a trend towards the Rust programming language, reflecting the changing direction of talent and skills. The real breakthrough lies in finding architectural concepts that can leverage environmental advantages. Current cybersecurity issues are severe, and microkernels may provide a fundamentally more resilient data processing environment. By employing diversity and redundancy in design, inspired by the evolutionary mechanisms of nature, open-source projects can build a more secure and stable IT infrastructure.
The IT industry is like a shark; it must keep swimming forward to survive. Not all sharks need to engage in ram ventilation, and not all IT is in constant flux, but without innovation, the industry will stagnate and die.
Venture capitalists and well-funded tech companies feel this most acutely, which is why they swarm to cutting-edge technologies like sharks. However, mere innovation is not enough. Blockchain is very clever, but it has made the world a more perilous place, causing many to lose substantial amounts of money. The trajectory of generative AI from a wait-and-see phase could also go in any direction. True transformation—innovation that can persist and become part of a better path forward—should not, and indeed must not, make its inventors billionaires.
A more productive future filter can be gleaned from nature. Evolution through selection and inheritance, this universal biological theory, serves as a living proof of successful innovation and a pathological library of failed innovations. It is simple enough to be a tautology—survival of the fittest—yet complex enough to be breathtaking in its interplay of environmental factors, selective pressures, change, and stability.
Applying this analysis of selective pressures to the specific phenomenon of species differentiation around the Linux kernel reveals clear pressures from external environmental changes: new and updated CPU architectures, evolving security models, performance expectations in different use cases, and various energy efficiency requirements. All of these drive changes in the kernel.
Other factors are more subtle, with distinctly different dynamic characteristics: the commercial and personal motivations involved, the available talent pool driving change vectors, the power exchange between closed and open systems, the accuracy of perceptions, and the anatomy of persuasion. These are both social and technical.
Applying this model of selective pressure to an overview of three microkernels in the Linux environment shows that feasible mutations tend to lean towards Rust—the combination of personal motivation and skills is shifting in that direction. Rust itself has evolved in response to these selective pressures, thus favoring entities that can more easily embrace it.
What can truly make an impact is when an architectural concept provides a pathway to leverage environmental characteristics that are unavailable or even toxic to existing organisms. The evolution of photosynthesis provided unprecedented energy to cyanobacteria, while the atmospheric oxygen produced was toxic to many other species. The evolution of metabolism that utilized oxygen completed a planetary reset event.
The IT equivalent is addressing security issues, which are currently so severe that in the UK, there have been proposals to adopt COVID-level response measures to maintain the survival of the entire supply chain after a single incident. Information theft, system infiltration, and ransomware are increasingly seen as inevitable, with corporate fatalism leaning towards survivalism. In an increasingly adversarial world, this is unsustainable.
Microkernels may seem unrelated to ransomware, but stepping back to ask when a mutation might lead to an inherently resilient data processing environment is crucial. Nature employs diversity and redundancy at all levels, from ecosystems to immune systems, to enhance resilience. Imagine making these the primary design features of a stack. In avionics and other safety-critical systems, diversity and redundancy mean that parallel independent systems perform the same tasks in different ways while cross-checking each other.
Building a microkernel that supports these low-level systems while simultaneously fulfilling functional IT tasks sounds like a very valuable experiment. Running another microkernel on different hardware is also important. These can begin as only open-source can, as stepwise modifications of existing code and ideas, turning small proof-of-concept builds into running systems, and encouraging others to start their own paths.
Nature has created and tested evolution for billions of years, resulting in magnificent phenomena. But we have goals and a sense of mission, which natural selection completely lacks. Artificial selection can achieve what nature may never accomplish in just a few years. We can use the same tools as nature but build in the direction we need.
We absolutely do not need a $300 billion fantasy AI supercloud to pursue the creation of the world’s first half-trillionaire. We absolutely do not need such a fragile IT infrastructure that allows a few unknown individuals to disrupt entire industries at will.
Understanding the big picture behind the small evolutions in open source can lead us to where we need to go. Evolving to live on dry land, away from sharks. All the sharks.
Q&A
Q1: Why is it said that the Linux kernel faces selective pressures?
A: The Linux kernel faces multiple pressures from the external environment, including new CPU architectures, evolving security models, performance requirements in different scenarios, and various forms of energy efficiency demands. Additionally, social-technical factors such as commercial motivations, changes in the talent pool, and power exchanges between open and closed systems also play a role.
Q2: How do microkernels address cybersecurity issues?
A: Microkernels can provide an inherently resilient data processing environment. Through diversity and redundancy in design, similar to parallel independent systems in avionics, they perform the same tasks in different ways while cross-checking each other, thereby enhancing security at the foundational system level.
Q3: What role does the Rust language play in the development of the Linux kernel?
A: Rust has become an important direction in the evolution of the Linux kernel, with personal motivations and skill combinations shifting towards Rust. Since Rust itself has evolved in response to selective pressures, these pressures will favor entities that can more easily embrace Rust, while mainstream Linux, due to the high institutional inertia of its maintainers, leaves opportunities for smaller, more agile projects.