technology sector: metabolising crisis

The technology sector increasingly finds itself in a strange position. It does not merely create new tools. It creates new forms of dependency, instability, and risk, then develops further technologies to manage the consequences. Cybersecurity exists because digital systems create vulnerability. AI safety exists because AI creates new forms of uncertainty. Trust infrastructure exists because trust has been degraded. The sector does not simply generate solutions. Increasingly, it metabolises its own waste products.

This creates a deeper problem. The same institutions driving technological acceleration remain dependent upon the social systems being destabilised by that acceleration. Wealth depends upon markets. Markets depend upon trust. Trust depends upon institutions. Institutions depend upon legitimacy. Legitimacy depends upon populations that believe the system works for them at least some of the time. Undermine enough of those foundations and the environment that produced technological success begins to weaken beneath the feet of those who benefited from it.

The concern around AI-enabled cybercrime sits squarely within this pattern. If offensive capability becomes dramatically cheaper and more scalable than defensive capability, then many assumptions underpinning modern society become unstable. Banking, identity, contracts, insurance, communications, governance, and trust itself all depend upon the practical difficulty of attacking systems. The issue is not that every account gets hacked tomorrow. The issue is that the balance may shift. A handful of actors with sufficiently capable systems could impose costs on millions of people faster than institutions can respond.

Most contemporary explanations begin with parts and attempt to assemble wholes from them. Individuals produce markets. Markets produce institutions. Institutions produce societies. Components produce systems. The analysis proceeds from local mechanisms upward, hoping that sufficient detail eventually explains the larger structure. The perspective here moves in the opposite direction. It begins by observing that wholes already exist. Societies exist. Economies exist. Ecologies exist. Languages exist. Civilisations exist. The question is not how isolated components somehow manufacture unity, but how unity constrains and organises the behaviour of components in the first place.

From that perspective, the technology sector is not external to the systems it influences. It does not simply create complex distributed risks and then offer tools to manage them. Increasingly, it metabolises its own waste products: dependency, volatility, insecurity, attention collapse, institutional disruption, and the structural problems generated by earlier layers of technological acceleration. Cybersecurity becomes a market because digital systems create vulnerability. AI safety becomes a market because AI systems create new forms of instability. Trust infrastructure becomes a market because trust has been degraded. Volatility itself becomes economically productive, metabolised into dependency, and dependency back into revenue.

Responsibility, then, is not merely a moral concern added after the fact. It is structural. A recurring pattern of modern technological development is that rewards are concentrated while consequences are distributed. New platforms emerge, investors profit, founders accumulate influence, and society is encouraged to celebrate innovation. Then the secondary effects arrive: misinformation, addiction, institutional erosion, labour disruption, surveillance, cybercrime, social fragmentation, and market concentration. At that point responsibility often becomes strangely difficult to locate. The gains are attached to identifiable actors. The costs become everyone else’s problem.

This is not simply a failure of individual corporations. It is a broader feature of the contemporary economic environment. Competitive pressure rewards acceleration more reliably than caution. Markets reward growth more reliably than stewardship. Organisations that pause to consider long-term consequences are frequently outcompeted by organisations willing to defer those considerations until later. The result is a civilisation that has become extraordinarily effective at generating capability while remaining comparatively poor at maintaining responsibility proportional to that capability.

There is also a deeper philosophical problem here. Classical reasoning often begins by separating things: A is A, A is not B. Yet the ability to distinguish A from B already presupposes a field within which both appear together. Difference requires relation. Separation quietly assumes connection. Unity arrives before distinction, even though distinction is what becomes visible. Much of modern technological culture forgets this. It treats systems as collections of exploitable parts, then seems surprised when the larger structures upon which those parts depend begin to deform.

The rush toward AI resembles a gold rush more than a planned civilisational transition. Millions of people are retraining, learning code, building agents, chasing investment, chasing jobs, and chasing relevance. The pressure is immense because nobody wants to be left behind. Yet much of the effort is occurring without a clear collective understanding of where the process leads. Everyone is accelerating. Far fewer are steering.

This does not mean AI is going to kill us all. That framing is too simple, and probably too flattering to the machines. The danger is not intelligence floating above history like a demon. The danger is humans with AI, which is simply humans with tools again, only faster, cheaper, more scalable, and more opaque. Atom bombs do not decide to kill people. The dickheads who choose to use them do. AI is not outside that pattern. It amplifies agency, including the agency of fools, criminals, bureaucrats, ideologues, executives, opportunists, and states.

Whether one calls this a singularity is almost beside the point. The popular image of a singularity was always a future event: a threshold beyond which change becomes difficult to predict. The more interesting observation is that many of the characteristic features are already present. Institutions struggle to adapt. Information moves faster than governance. Economic structures are being reconfigured faster than political structures. Human cognition is increasingly coupled to machine systems whose behaviour is only partially understood. The discontinuity may not arrive as a single moment. It may be the condition we are already living through.

The question is not whether AI becomes more powerful. It almost certainly will. The question is whether societies can remain coherent while their primary mechanisms of communication, coordination, production, persuasion, surveillance, and decision-making are all being transformed simultaneously.

AI is not going to kill us. Humans have always been perfectly capable of causing trouble without machine intelligence. The danger is humans with AI, just as the danger was never the atom bomb but the people willing to use one. The technology amplifies agency. It does not replace it.

The deeper risk is that we become so focused on the power of the tools that we forget the conditions that make those tools useful in the first place. A civilisation can survive imperfect technology. It cannot indefinitely consume the trust, legitimacy, stability, and social coherence upon which everything else depends.