Essay · Technological History · Futures

The Pace
of Everything

On how the speed of change has always exceeded our capacity to absorb it, and why this time may be structurally different.

EvolvingSoftware.com 2026 Long Read · ~12 min
Scroll
Every technology that has ever disrupted the world was, at first, absorbed slowly and then all at once, until the next one arrived before the absorption was complete.
Preface On the nature of technological time

There is a question beneath every conversation about artificial intelligence, about Evolving Software, about the computational systems that are beginning, structurally and not metaphorically, to modify their own trajectories. The question is not whether change is coming. The question is whether the pace of that change exceeds the pace at which human civilisation can adapt to it.

History offers something rare and useful here: not comfort, exactly, but pattern. Humanity has navigated transformative technological disruption before. The written word. The printing press. The steam engine. The telegraph. The internet. Each of these reshaped not merely what people did, but how they thought, how they organised, how they understood themselves and their place in a social order. And each time, in the rearview mirror, the adaptation looks almost inevitable.

But the rearview mirror is a deceptive instrument. It smooths the chaos into a trajectory. It hides the grief of those who could not adapt quickly enough: the artisans, the clerks, the navigators, the switchboard operators. It makes the disruption look like progress. Sometimes it was. Sometimes it was both at once.

What follows is an attempt to read the pattern honestly, to locate the present moment inside a long history of disruption, and to hold two truths simultaneously: that the human race has proven remarkable in its adaptability, and that the architecture of current change may be unlike anything that has preceded it in one critical dimension: speed.

The Long Arc: Four Eras of Transformative Change
c. 3200 BCE – c. 1450 CE
The Written Word
Cuneiform, papyrus, the codex, and eventually moveable type. Knowledge became portable, repeatable, and eventually scalable. The absorption time: centuries to millennia. Each civilisation had time to grow into its archive.
c. 1760 – c. 1900
Industrialisation
Steam, coal, and the factory system. The compression of space and time through rail and telegraph. Absorption time: three to four generations. Brutal for those who lived through the transition; generational prosperity for those who came after.
c. 1990 – c. 2015
The Internet Era
The collapse of information geography. A single connected layer over the entire planet. Absorption time: one generation, fast enough to strand many, slow enough to enable meaningful adaptation and new institutional forms.
c. 2015 – Present
The Emergence Era
Machine learning, large-scale AI, Evolving Software architectures. Absorption time: unknown. For the first time, the rate of change may be exceeding our institutional capacity to model it, let alone govern it.
I The shape of past disruption

What the Past Actually Shows Us

There is a version of technological history that reads as a clean ascent: each innovation lifts productivity, productivity lifts living standards, living standards lift life expectancy and opportunity. On aggregate, at civilisational scale, this is broadly true. The introduction of the printing press eventually produced mass literacy. The industrial revolution eventually produced the eight-hour workday. The internet eventually produced more equitable access to information than any library system could provide.

The key word is eventually. The mechanism is correct but the temporal distribution matters enormously. Benefits tend to arrive slowly and unevenly, clustering first with those who have capital, geography, or the specific skills the new technology rewards. Disruption tends to arrive faster and more broadly, cutting across entire industries, geographies, and ways of life before the compensating structures (the new schools, the new regulations, the new economic roles) have had time to form.

The handloom weavers of northern England during the 1810s were not wrong to resist the power loom. The disruption to their livelihoods was real, immediate, and devastating, while the long-run prosperity it would eventually enable was abstract, distant, and distributed to people not yet born. The Luddites were not anti-technology. They were pro-survival. The lesson history draws from their defeat is not that they were wrong about the pain. It is that they could not stop the structural inevitability of the change.

This asymmetry, real costs now and diffuse benefits later, is not a flaw in the historical record. It is the consistent structural signature of major technological transitions. Understanding it honestly is the prerequisite for navigating what is ahead.

The most dangerous assumption in any era of technological transition is that because it worked out before, it will work out again, in roughly the same way, over roughly the same timeframe.

EvolvingSoftware.com, 2026
II What makes now structurally different

The Compression Problem

The framework described on this site identifies temporal compression as its sixth structural layer of emerging complexity, and it is the layer most immediately relevant to any discussion of societal adaptation. The same structural dynamics, operating at radically different clock speeds, produce qualitatively different outcomes.

Consider the printing press. Gutenberg's invention was completed around 1440. By 1500, approximately 8 million books had been printed across Europe. By 1600, that number had grown to around 200 million. But the social and political consequences (the Reformation, the Scientific Revolution, the emergence of national languages as literary standards) unfolded over the following two centuries. The technology was fast. The absorption was generational.

Now consider the trajectory of large language models. GPT-2 was released in 2019 with enough capability to write coherent paragraphs. By 2023, GPT-4 was passing bar exams and medical licensing tests. By 2025, autonomous software systems were beginning to write, test, and deploy other software without human instruction at each step. The capability curve compressed what was previously a multi-generational transition into less than a decade.

This is not merely a faster version of what has happened before. It is a different regime. Institutional adaptation, encompassing regulatory frameworks, educational curricula, labour market adjustments, legal structures and social norms, operates on timescales measured in years to decades. When technology moves faster than this, the gap between capability and governance does not simply widen. It becomes structurally unstable.

The Three Paths Ahead
Sensitivity Analysis: A Range of Outcomes
〰
The Absorptive Path
Gradual Integration

In this scenario, the pace of change, while unprecedented, remains within the range of institutional adaptation. Regulatory frameworks develop alongside capability. Educational systems evolve to meet new requirements. New economic roles emerge at sufficient scale to absorb those displaced. The productivity gains are large and, over time, broadly distributed.

This is not the comfortable scenario. It still involves significant disruption, real losses, and genuine hardship for those caught in the transition. But it has historical precedent. It is the path that has, broadly, characterised prior transitions, simply compressed into a tighter timeframe. It requires coordination between technology developers, governments, and institutions that is rare but not impossible.

△
The Uneven Path
Asymmetric Distribution

In this scenario, arguably the most probable of the three, the productivity gains are real and substantial, but profoundly asymmetric in their distribution. The capability of Evolving Software concentrates value in the hands of those with access to capital, compute, and technical expertise, while hollowing out the economic roles that have historically provided stable livelihoods for large portions of the population.

This is not a new dynamic. The industrial revolution produced this pattern. The internet produced this pattern. But the magnitude may be different this time, and critically the buffer time for social adaptation may be insufficient. When transitions were slower, welfare states, unions, and regulatory frameworks had time to form. When transitions are fast, that buffer shrinks toward zero.

∅
The Rupture Path
Black Swan Event

In this scenario, one or more events occur that are not merely disruptive but structurally destabilising: events that invalidate the operating assumptions of institutions, economies, and social contracts simultaneously. Not a faster version of historical disruption, but something qualitatively different: a system that modifies the rules of its own modification before human governance can respond.

This is the scenario for which history provides the least guidance, because history has no prior examples of a technology with the capacity to autonomously rewrite its own architecture at scale. The black swan, by definition, cannot be modelled from within the paradigm it will rupture. What can be said is that the preconditions for such an event are becoming more structurally plausible with each passing year.

The Black Swan Record
Thirty Years of Rupture

Since 1990, the frequency of low-probability, high-impact events that invalidate prior operating assumptions at civilisational scale has increased measurably. This is not coincidence. Interconnected, interdependent, accelerating systems are structurally more brittle than isolated, slower ones.

c. 1985–1995
Personal Computing Goes Mass Market
The democratisation of computational power, placing programmable machines in homes and offices worldwide and seeding an entirely new economic layer.
c. 1995–2000
Internet Adoption
A billion-person communications network deployed in under a decade, collapsing the geography of information and commerce in ways no institution had modelled.
c. 2000–2015
Digital Commerce Displaces Bricks and Mortar
The structural hollowing-out of retail, travel, media and financial services by platforms operating without the cost base of the incumbents they replaced.
c. 1990–2010
International Travel and Global Connectedness
Passenger air travel tripling in two decades, compressing cultural distance and creating the biological and informational interdependence that makes all subsequent shocks faster and wider.
2008
Global Financial Crisis
Systemic interdependence collapsing faster than any model predicted.
c. 2007–2015
Streaming Reshapes Entertainment Distribution
Broadband bandwidth enabled the on-demand delivery of film, music and television, dismantling century-old distribution models and concentrating cultural gatekeeping into a handful of global platforms.
c. 2008–2020
The Collapse of Mainstream Print and the Rise of Social Media
Advertising revenue migrated to platforms overnight, hollowing out newsrooms globally. The public sphere, once distributed across thousands of independent editorial voices, concentrated into algorithmic feeds controlled by a small number of private companies.
2020
COVID-19
A global system halt that compressed a decade of digital transition into 18 months.
2022
GPT Inflection
The moment AI moved from specialist tool to general capability at consumer scale.
2024–25
Agentic Systems
Software that writes, tests, and deploys other software without human instruction.
III Human adaptability and its structural limits

The Enduring Case for Human Adaptability

It would be a mistake to conclude from the above that pessimism is the correct response. The record of human adaptability is genuinely extraordinary. Homo sapiens survived the Toba catastrophe, a volcanic eruption approximately 74,000 years ago that may have reduced the global human population to as few as ten thousand individuals. We navigated the Black Death, which killed between a third and a half of Europe's population in four years. We absorbed the industrial revolution, two world wars, the invention of nuclear weapons, and the complete restructuring of the global economy, not once but three times in the span of a single century.

The mechanism of this adaptability is worth understanding precisely, because it is also its limitation. Humans adapt culturally more than biologically. We build institutions, laws, norms, educational systems, welfare structures and professional frameworks, that encode the lessons of prior disruptions and create scaffolding for navigating future ones. This is powerful. It is also slow. Cultural adaptation operates on generational timescales. And the critical question of our present moment is whether the acceleration of capability is outpacing the formation of adaptive institutional scaffolding.

There are reasons for genuine optimism here. The global response to COVID-19, for all its failures, demonstrated remarkable institutional agility: vaccine development timelines that would have been considered impossible a decade earlier were achieved in months. Open-source AI safety research has grown from a niche concern to a well-funded global field in less than five years. The institutions are moving. The question is whether they are moving fast enough, and whether the Evolving Software systems themselves, as they begin to modify their own trajectories, will remain within the governance envelope long enough for that adaptation to complete.

IV The connection to Evolving Software

Where Evolving Software Enters the Frame

The framework described on this site, the seven structural layers by which systems develop the capacity to participate in their own transformation, is not merely a description of what AI might eventually become. It is a map of what is already structurally forming, layer by layer, in the computational systems being built and deployed today.

Replication. Variation. Feedback-guided direction. These are not future capabilities. They are present architectural features of systems already deployed at scale. What remains nascent, but advancing rapidly, is the cascading interdependence layer: the point at which systems begin to influence each other's trajectories in ways that produce emergent coordination without any single agent holding the design.

This is precisely the structural condition that creates the gap between optimism and concern. If Evolving Software develops through clearly legible stages, with human oversight viable at each transition, then the absorptive scenario becomes plausible. If the cascading interdependence layer develops faster than legibility allows, and if the emergence is structurally opaque as well as structurally inevitable, then governance faces not a technical problem but an epistemological one: how do you govern a system whose behaviour cannot be fully anticipated or modelled?

This is the question that sits beneath every conversation about AGI timelines, alignment research, and AI safety. And it is a question that demands the kind of structural analysis this framework provides, not as speculation, but as a map of the architecture. The trajectory was always forming. The responsible act is to see it clearly.

On Living at the Inflection

Every generation believes it lives at a unique historical moment. Most are wrong, or rather, the uniqueness they perceive is real but not singular. What is genuinely different about the current moment is not the presence of disruption, but the structure of it: for the first time in history, the disruption is coming from a process that can, in principle, modify its own disruption rate.

That is the meaning of Evolving Software. Not that machines will become conscious. Not that AI will replace all human labour. But that we are approaching, and in some domains have already passed, the structural threshold where software ceases to merely execute instructions and begins to modify trajectories. That shift, combined with temporal compression, combined with cascading interdependence between systems, creates a set of conditions that are genuinely without historical precedent in their specific architecture, even if the broad shape of technological disruption is familiar.

The honest position is not optimism or pessimism. It is clarity. The pattern of past transitions gives reason to believe that human adaptability, at civilisational scale, can meet the challenge. The structure of the current transition gives reason to believe that the margin for error is smaller than it has ever been, and that the cost of complacency, of assuming the transition will unfold on historical timescales and in historical forms, is higher than it has ever been.

The middle path is available. It requires seeing the architecture clearly and acting on that clarity while the window of legibility remains open.

The framework exists not to alarm, but to illuminate. What is seen clearly can be navigated. What is navigated, however imperfectly, can become something better than what preceded it. That has always been the story. The question is whether we are moving fast enough to make it true again.