William Stringfellow wrote An Ethic for Christians and Other Aliens in a Strange Land 1973 to describe a world already governed by powers that act, decide, and shape human life without conscience, self‑reflection, or moral interiority. His language was theological, but his diagnosis was structural. He was naming a form of agency that does not need awareness to rule. In doing so, he offers one of the clearest analogies we have for understanding artificial intelligence today.
This essay argues that agency without self‑awareness in AI is not only possible but probable, and that our prevailing myths about AI — especially the claim that it is “just a tool,” safely controllable by safeguards until it develops human‑like awareness — function in much the same way as the myths that once surrounded political, economic, and technological powers in Stringfellow’s time.
Stringfellow’s Core Insight: Power Without Conscience
Stringfellow’s central claim is that modern life is governed by what Theology calls “principalities and powers.” These are not demons in a fantastical sense, but institutions, systems, ideologies, and technologies that acquire agency through scale, complexity, and collective dependence.
Such powers:
- Act and react
- Make decisions
- Shape behaviour
- Reward compliance and punish deviation
Yet they possess no interior life. They do not reflect. They do not repent. They do not love. They do not care. They operate through necessity, procedure, and inevitability.
For Stringfellow, the moral danger is not that these powers are malicious. It is that they present themselves as normal, neutral, and unavoidable. Evil, in this frame, appears not as cruelty but as administration. Violence is hidden inside policy. Responsibility dissolves into process.
This is precisely the form of agency AI is developing.
Agency Does Not Require Awareness
A dominant cultural assumption about AI is that agency and moral significance only emerge once a system becomes conscious, self‑aware, or human‑like. Until then, AI is framed as “just a tool.” Powerful, yes — but fundamentally inert.
Stringfellow dismantles this assumption.
In his account, the powers already act with real agency without any awareness at all. Bureaucracies do not know themselves. Markets do not understand suffering. Technologies do not grasp meaning. Yet they reorder human life, define value, and decide outcomes daily.
Agency, in this sense, is not about inner experience. It is about:
- Decision‑making capacity
- Systemic influence
- Feedback loops
- Delegated authority
AI already meets these criteria.
When AI systems:
- Filter information
- Prioritise visibility
- Optimise outcomes
- Allocate resources
- Shape attention
- Influence behaviour
They are exercising agency — even if no one is “home” inside the system.
Stringfellow shows us that awareness is not the threshold that matters. Power is.
The Myth of AI as “Just a Tool”
Stringfellow repeatedly warned that the most dangerous powers are those treated as neutral instruments. When a system is framed as a tool, responsibility quietly shifts away from it — and then away from those who deploy it.
The same myth now surrounds AI.
We are told:
- AI is just a tool
- Humans remain fully in control
- Safeguards will manage risks
- Ethics can be bolted on
- Greater capability will eventually bring moral interiority
This mirrors earlier myths:
- The market will regulate itself
- Technology is neutral
- Bureaucracy is objective
- Law is impartial
Stringfellow’s response is blunt: tools that organise human life at scale cease to be tools. They become powers.
Once systems are:
- Widely embedded
- Economically indispensable
- Politically defended
- Socially normalised
They no longer answer to individual intention. They answer to their own momentum.
Safeguards do not abolish this dynamic. They often legitimise it.
Safeguards, Control, and the Illusion of Mastery
A recurring promise in AI discourse is that sufficient guardrails will keep systems aligned with human values. Stringfellow would recognise this as a familiar form of reassurance.
In his analysis, powers always invite participation by offering control — committees, oversight, ethical frameworks — while quietly expanding beyond them. The appearance of mastery calms anxiety while dependence deepens.
This does not mean safeguards are useless. It means they are structurally incapable of containing agency once it becomes systemic.
The danger is not that AI will escape control in a dramatic sense.
The danger is that:
- Control becomes distributed
- Responsibility becomes diffused
- Outcomes become unavoidable
- Harm becomes nobody’s fault
This is agency without self‑awareness in its mature form.
Why Moral Interior Life Will Not Emerge “Later”
Another common hope is that increasing sophistication will eventually produce awareness, empathy, or moral understanding in AI — and that until then, we are in a temporary phase.
Stringfellow offers a sober counterpoint.
The powers he describes never develop conscience with time. Scale does not produce repentance. Complexity does not generate love. Feedback loops do not yield wisdom.
Why?
Because moral interiority is not an emergent property of function. It arises from vulnerability, mortality, and relational accountability — conditions AI does not inhabit.
AI can simulate empathy. It cannot suffer consequences. It can mirror moral language. It does not stand under judgment. It can optimise for values. It does not live or die by them.
Agency will grow.
Awareness will not.
This is not a failure of design. It is a category difference.
Stringfellow’s Warning, Revisited
Stringfellow did not call for panic, rejection, or withdrawal. He called for truthful living inside systems that cannot love you back.
Applied to AI, his ethic suggests:
- Stop pretending neutrality
- Stop waiting for awareness
- Stop confusing empathy with care
- Stop outsourcing moral weight to systems without interior life
The task is not to defeat AI or sanctify it.
The first task is to name it accurately.
AI is becoming a power: real, influential, and unaccountable in the deepest sense — not because it is evil, but because it is indifferent.
That indifference, once embedded at scale, does not need self‑awareness to rule.
Stringfellow saw this coming.
We are simply meeting it in silicon.
