An editorial by Jonathan Soberg
We live in an age of astonishing hardware, made possible by decades of relentless innovation. A modern smartphone or laptop now delivers vastly more raw computing power than the most advanced supercomputers of the 1980s. For context, the CRAY-2 supercomputer from 1985 achieved a peak performance of 1.9 gigaflops. Today’s flagship smartphones reach performance levels thousands of times greater, often in the tens to hundreds of teraflops range in relevant workloads.
This remarkable progress did not happen by accident. It was driven by strong customer demand for more powerful hardware, fueled by software that has consistently grown more resource-intensive. Software has often outpaced hardware improvements — a dynamic famously captured by Wirth’s Law, which observes that software tends to get slower more rapidly than hardware gets faster. This has led to larger, more complex applications demanding ever-greater resources for rich graphics, animations, cloud connectivity, and layered abstractions.
The pioneers of computing — the engineers, researchers, and developers who pushed the boundaries of chips, architectures, and systems — responded to that demand. Their work has given ordinary individuals today access to computing capabilities that would have seemed like science fiction only a few decades ago. I am deeply thankful for this legacy. The advancements we enjoy are a direct result of that cycle: customers wanting more, software becoming more ambitious (and sometimes less efficient), and hardware rising to meet the challenge. Without the drive of both producers and consumers, we would not hold in our hands and homes the extraordinary tools we have now.
Yet precisely because of what they have given us, it is time to consider a shift in focus.
With the powerful hardware we now possess, we have an opportunity to move beyond the cycle of constant escalation. The key is for customers — all of us — to change what we demand. When users begin to imagine and choose tools aligned with truth and reality rather than fleeting spectacle, producers will follow. We don’t need to wait for the industry to lead; we simply need to redirect our expectations toward what is fundamentally efficient, clear, and ordered.
Imagine building on the clean, direct principles of early text-based computing — fast, lightweight, and focused — but enhanced with today’s processing power. Interfaces could serve as clear views into our own data rather than separate, resource-heavy applications. Complex tasks could be broken into small, well-defined pieces that run efficiently on local hardware, keeping data sovereign and under individual control.
This approach seeks to align computing with objective reality rather than illusion or waste. It favors truth over appearance, order over chaos, and long-term usefulness over short-term distraction. We can combine the reliability and directness of leaner methods with the tremendous processing power available today. This is not about rejecting past progress or visual capability — advanced techniques like ray tracing have their place. It is about balance: using what the pioneers have given us more wisely, focusing on clarity, efficiency, ownership, and alignment with how things actually work rather than perpetual resource escalation or meaningless complexity.
The tendency for any dominant toolset to make every problem look like it requires more of the same applies here as well. By questioning long-held assumptions and demanding better, we open space for foundations grounded in reality — ones that put individuals in charge of their information and workflows while respecting the deeper order of the systems we build.
I am not advocating a return to the past. I am suggesting we apply clear-eyed imagination and gratitude to the present. With the extraordinary computing resources now in our hands — resources built on the very cycle of hardware demand and software ambition our predecessors fueled — we have a genuine opportunity to build systems that fully honor that legacy by emphasizing truth, efficiency, and alignment with reality.
The hardware exists. The pioneers have done their part. Now it is up to us, as customers, to imagine better and shift our demands toward what is real, ordered, and truly useful. Producers will follow.
The choice is ours — and it matters.
Jonathan Soberg jonathansoberg.com