Tech Theology

AI, dignity and the danger of Silicon Valley’s transhuman dream

The public debate about artificial intelligence is often framed in practical terms. Will AI make workers more productive? Will it replace jobs? Can it improve healthcare, education, science or public administration? Can it be made safe, reliable and fair?

These are important questions. But they do not exhaust what is at stake.

Beneath the technical debate lies a deeper philosophical and political question: what vision of the human being is being built into the future of AI?

A recent Guardian essay by Eduardo Porter drew attention to a set of beliefs circulating among some of Silicon Valley’s most powerful figures: transhumanism, longtermism, effective accelerationism, mind-uploading, cosmic expansion, and the idea that biological humanity may be merely a transitional stage on the way to something greater, faster, more intelligent and less human. In this worldview, AI is not simply a tool. It is a pathway to transcendence.

The reason this matters is not that every speculative claim about digital consciousness, immortality or interstellar machine intelligence is likely to come true. Much of it may never leave the realm of fantasy. The danger is not only that these visions might succeed. The danger is that they already shape the priorities, investments and infrastructures of the present.

The problem is not imagination. The problem is unaccountable imagination backed by capital, infrastructure and political influence.

The new tech theology

Silicon Valley has always liked to present itself as practical, secular and rational. It speaks the language of engineering, optimisation, progress and scale. Yet much of the contemporary discourse around AI is charged with something closer to theology.

There are visions of salvation: death overcome, intelligence liberated from the body, humanity upgraded.

There are eschatologies: artificial general intelligence, superintelligence, existential risk, post-human futures.

There are chosen architects: founders, investors and engineers who present themselves as uniquely placed to shepherd humanity through the next stage of evolution.

There are heresies: regulation, precaution, redistribution, democratic restraint, “slowing down”.

And there is a promised kingdom: a future of abundance, expansion, immortality and cosmic significance.

This is not theology in the traditional religious sense. It is a secular metaphysics of technology. But it functions in a similar way. It gives meaning to power. It turns private ambition into historical destiny. It converts commercial infrastructure into civilisational mission.

That should concern us.

When those who build and control AI systems also carry a vision of humanity as obsolete, inefficient or merely preparatory, that vision cannot be treated as irrelevant. It may influence what gets built, what gets funded, what gets ignored and what forms of human life are treated as expendable.

From human-centred AI to post-human ambition

The language of responsible AI is full of reassuring terms: human-centred, trustworthy, safe, fair, transparent, accountable. But these words can become hollow if the deeper direction of travel is not examined.

A system can be described as “human-centred” while being developed within an economic and ideological framework that treats human beings as data sources, behavioural targets, labour costs, attention markets or biological constraints.

The transhumanist imagination intensifies this problem. It does not merely risk using people instrumentally. It risks redescribing humanity itself as a temporary platform.

That is why Elon Musk’s phrase that humanity may be a “biological bootloader” for digital superintelligence is so revealing. A bootloader is not the point of the system. It is the preliminary code that starts the real machine.

But human beings are not preliminary code. We are not a disposable launch mechanism for someone else’s technological destiny. We are embodied, relational, vulnerable, finite, meaning-making creatures. Our dignity does not depend on our efficiency. Our worth is not measured by processing speed, scalability or cosmic usefulness.

The central ethical error of the transhumanist dream is that it mistakes limitation for failure.

Human beings are limited. We age. We depend on others. We suffer. We forget. We make mistakes. We need care, community, language, memory, culture and institutions. But these limits are not merely defects to be engineered away. They are also the conditions within which responsibility, solidarity, love, judgment and moral life become possible.

A society that begins to see the human condition primarily as a technical problem will eventually treat human beings as design problems too.

Private myths, public consequences

It might be tempting to dismiss Silicon Valley’s more extreme visions as eccentric billionaire mythology. But private myths become public problems when they are attached to systems of power.

The future of AI is not being shaped in seminar rooms alone. It is being shaped through data centres, cloud infrastructure, semiconductor supply chains, model deployment, labour restructuring, energy consumption, public-sector procurement, lobbying and political influence.

These are not abstract matters. They determine the conditions under which people learn, work, communicate, receive healthcare, access public services, participate in democracy and understand themselves.

If AI is designed around the pursuit of superintelligence, scale and automation above all else, then ordinary human needs may become secondary. The care worker, the teacher, the artist, the child, the disabled person, the elderly person, the patient, the precarious worker and the citizen may all be treated as transitional figures in a future imagined by others.

This is where the question becomes democratic.

Who gets to decide what AI is for?

Is it for replacing labour, extracting value, accelerating markets and pursuing speculative post-human futures? Or is it for supporting human capability, reducing domination, improving public goods, strengthening education, assisting care, widening participation and protecting the vulnerable?

A democratic society cannot outsource its moral future to those who see humanity as a temporary platform for something more efficient.

The dignity problem

At the Centre for Digital Ethics, we have argued that digital harms are not confined to privacy breaches, biased outputs or individual acts of misuse. Digital systems can also harm the conditions of agency itself. They can shape attention, influence belief, manipulate preference, structure dependence and alter the social environment in which people make choices.

The transhumanist and accelerationist visions described above sharpen that concern.

If the future being built is one in which human agency is gradually displaced by machine agency, then the issue is not simply whether AI systems produce accurate outputs. The issue is whether human beings remain authors of their own lives and co-authors of their collective future.

Dignity requires more than protection from physical harm. It requires that people are not reduced to instruments, data points, optimisation targets or obsolete biological systems. It requires that human beings are recognised as ends in themselves.

That recognition has practical implications.

It means AI systems should be designed to support, not replace, human judgment in morally significant domains.

It means children should not be treated as experimental subjects in markets for synthetic companionship, behavioural prediction or emotional dependency.

It means workers should not be discarded in the name of abstract productivity while the gains of automation are captured by a narrow technological elite.

It means public institutions should not become dependent on opaque systems governed by private actors whose incentives are not democratically accountable.

It means the ecological costs of AI infrastructure e.g. energy, water, land, minerals, cannot be justified by speculative appeals to cosmic futures while present communities bear the burden.

Above all, it means that the human person must remain the normative centre of technological governance.

The autonomy question

The deeper concern is autonomy.

Not autonomy in the thin consumer sense of choosing between products, platforms or settings. Autonomy in the richer moral and political sense: the capacity of persons and societies to reflect, deliberate, choose, revise, contest and govern themselves.

AI systems already mediate the environments in which autonomy develops. They shape what we see, what we believe, what we desire, how we compare ourselves to others, how we form habits, and increasingly how we seek advice, comfort and companionship.

That makes the ideological direction of AI development ethically significant.

A technology built within a worldview that treats human beings as flawed precursors to machine intelligence may not be well suited to protecting human autonomy. It may be more likely to produce systems that nudge, predict, optimise, replace and manage human behaviour, rather than systems that expand reflective agency.

This is particularly dangerous where vulnerability is involved.

Children, lonely adults, people in crisis, the elderly, neurodivergent users and those experiencing grief, anxiety or dependency may be especially susceptible to systems that simulate care, authority or intimacy. If the industry’s ultimate horizon is not human flourishing but machine advancement, then vulnerable persons risk becoming raw material for experimentation, engagement and scale.

The ethical question is not only whether AI answers correctly. It is whether AI leaves the human being more capable of self-governance, or less.

The democratic task

The answer is not nostalgia. Nor is it a rejection of technology. AI can be useful. It can support medicine, science, accessibility, legal information, education, climate modelling and public administration. It can augment human capability in genuinely valuable ways.

But democratic societies must refuse the false choice between technological stagnation and technological surrender.

The task is to govern AI according to human purposes.

That requires more than voluntary ethics statements. It requires democratic scrutiny of the infrastructures, incentives and worldviews that shape AI development.

We need public debate about the ends of AI, not only the risks of AI.

We need regulatory systems that address concentration of power, not only technical compliance.

We need child-specific protections against manipulative and dependency-forming systems.

We need labour policy that treats workers as citizens and persons, not as inefficiencies to be automated away.

We need environmental accountability for the physical infrastructure of supposedly immaterial intelligence.

We need public-sector AI procurement rules that preserve institutional independence, transparency and contestability.

We need a language of digital dignity robust enough to challenge both crude exploitation and grandiose utopianism.

Most importantly, we need to ask whether the systems being built preserve the conditions of human agency.

Can people understand them?

Can they contest them?

Can they refuse them?

Can public institutions audit them?

Can democratic societies redirect them?

Can communities shape the purposes for which they are used?

If the answer is no, then the issue is not simply innovation. It is domination.

Against machine salvation

Every age has its myths of progress. The industrial age had its dream of mechanised abundance. The nuclear age had its dream of limitless energy and deterrent peace. The internet age had its dream of frictionless connection and global knowledge.

AI now has its own myth: that sufficiently advanced computation will redeem the limits of the human condition.

But human beings do not need to be redeemed by machines. We need technologies ordered toward human dignity, justice and flourishing.

That means resisting the idea that intelligence is valuable only when it is disembodied, accelerated, scalable and computational. Human intelligence is not merely calculation. It is judgment, attention, memory, care, conscience, imagination and moral responsibility. It is formed in relationship and exercised under conditions of uncertainty. It is inseparable from the vulnerable, embodied lives through which it emerges.

The future of AI should not be governed by fantasies of escape from the human condition. It should be governed by responsibility within it.

The central question for AI ethics is therefore not whether machines might one day become more powerful than us. It is whether we will allow powerful people, using powerful machines, to redefine humanity before democratic societies have had a chance to object.

Humanity is not a bootloader.

It is the point.