From “Don’t Do It, Because I’m Watching You” to “Don’t Do It, Because You Understand What You Are Causing the Other”
The Transition from External Surveillance to Moral Autonomy in the Age of Artificial Intelligence
Abstract
This essay examines the transition from an ethics of external surveillance and coercion —“Don’t do it, because I’m watching you”— to an ethics of internal understanding and autonomy —“Don’t do it, because you now understand what you are causing the other.” The argument begins from the hypothesis that human beings possess deep dispositions toward reciprocity, empathy, cooperation, and mutual restraint that precede modern institutions of law, without implying that humanity once possessed a single, universal, and immutable “natural morality.” The discussion connects this problem with Aristotle’s conception of virtue and practical wisdom, Kant’s concept of autonomy, Emmanuel Levinas’s primacy of responsibility toward the Other, and Jürgen Habermas’s account of communicative rationality, discourse ethics, law, and democratic legitimacy. Particular attention is given to Artificial Intelligence: it may either strengthen systems of surveillance and behavioral control or contribute to the development of human moral understanding. The central argument is that technological progress should not be measured merely by its capacity to create more efficient mechanisms of compliance, but also by its capacity to strengthen human beings’ ability to recognize the Other and understand the consequences of their actions.
Keywords: natural morality, Golden Rule, moral autonomy, state, surveillance, Other, Artificial Intelligence, posthumanism, ethics of technology.
DIOTIMA
1. Introduction: Two Sentences, Two Moral Worlds
“Don’t do it, because I’m watching you.”
“Don’t do it, because you now understand what you are causing the other.”
Between these two sentences lies an entire philosophical horizon.
The first describes a human being who regulates conduct because there is an external gaze, a rule, a prohibition, and ultimately the possibility of punishment. The second describes a human being who has internalized the moral relationship with the other to such an extent that external surveillance becomes less necessary.
The difference is not merely a difference in the degree of compliance. It concerns the quality of human existence itself.
In the first case, the human being obeys.
In the second, the human being understands.
In the first, the law exists outside the individual.
In the second, the moral principle has become part of consciousness.
The question, therefore, is not whether societies need rules. They do. The deeper question is: to what extent should social order depend upon surveillance and external coercion, and to what extent can it depend upon internal moral self-regulation?
And, at an even deeper level: could Artificial Intelligence contribute to the transition from the human being who behaves properly because he is being watched to the human being who acts responsibly because he understands the Other?
2. The Hypothesis of Natural Morality and the Golden Rule
The idea of a “natural morality” must be approached with philosophical caution. There is insufficient historical evidence to establish that humanity once possessed a single, universal, and immutable moral condition. Human societies have differed substantially in their values, norms, and conceptions of good and evil.
A more modest and defensible proposition is possible: human sociality appears to include deep capacities for cooperation, reciprocity, mutual assistance, and sensitivity to harm inflicted upon others.
The so-called Golden Rule —“do not do to another what you would not want done to you”—can be understood as a condensed expression of this logic of reciprocity.
Yet the Golden Rule does not resolve every moral problem. The fact that I would tolerate an action does not mean that another person wishes to experience it. Morality therefore requires more than projecting oneself onto the other. It requires recognition of otherness.
This leads from the Golden Rule toward a deeper moral principle:
Do not act against another when you understand that your action violates that person’s freedom, dignity, or capacity to live as an autonomous person.
At this point morality ceases to be merely a rule of reciprocity and becomes a relationship of responsibility.
3. Aristotle: Morality as the Formation of Character
Aristotelian ethics offers an important counterpoint to the idea that the moral person is simply someone who knows the correct rules.
For Aristotle, virtue is not mere knowledge. It is a disposition of character, formed through habituation and practice. Ethical action also requires practical wisdom —phronesis— the capacity to perceive what a particular situation requires and to act accordingly.
This is particularly important for the present argument.
The moral person is not someone who waits each time to be informed about what is permitted and what is forbidden. The moral person is someone whose character has been cultivated sufficiently to recognize the appropriate course of action within concrete circumstances.
Aristotle therefore moves us from obedience to a rule toward the formation of a human being capable of acting well.
This distinction is crucial for Artificial Intelligence as well. If AI merely informs human beings about what is permitted, it remains a technology of normative compliance. If, however, it helps human beings understand the consequences of their actions, it may become something closer to a technology for strengthening moral judgment.
4. Kant: From Heteronomy to Autonomy
Kantian philosophy provides a second fundamental dimension.
Moral autonomy means that a human being is not merely directed by external commands but is capable of governing himself or herself through principles recognized as binding. Autonomy occupies a central place in Kantian ethics and is connected with the idea that morality cannot ultimately derive its authority from external coercion.
This distinction allows us to understand the two opening sentences more deeply.
“Don’t do it, because I’m watching you” is paradigmatically heteronomous.
“Don’t do it, because you understand that you cannot treat the other merely as a means” moves toward autonomy.
Moral maturity, therefore, is measured not only by whether a person does the right thing, but also by why that person does it.
A person who avoids crime solely because he fears punishment behaves lawfully. That fact alone does not establish moral autonomy.
5. Levinas: The Other Before the Rule
If Kant leads us toward autonomy, Emmanuel Levinas leads us toward the Other.
For Levinas, ethics does not begin with an abstract system of rules but with the encounter with another person and the responsibility generated by that encounter. The relation to the Other precedes, in an important sense, the theoretical construction of a complete moral system.
From this perspective, the second sentence acquires a new depth:
“Don’t do it, because you understand what you are causing the other.”
It does not merely say: “Do not violate a rule.”
It says: “See the human being standing before you.”
Ethics therefore moves from an impersonal rule toward a concrete relationship.
The word “see” acquires two radically different meanings.
In “I am watching you,” the gaze is power.
In “I see the other,” the gaze becomes recognition.
This may constitute one of the most important transformations human civilization can undertake.
6. The State: Necessary Institution or Exclusive Moral Authority?
From here emerges the political question.
Why should the state be the sole source of authority concerning what human beings ought to do?
The answer cannot simply be that the state creates morality itself. Morality precedes legislation both logically and culturally. Societies develop values, norms, and expectations, and some of these subsequently acquire legal form.
The state, therefore, need not be the source of morality. It is the institution possessing legitimate authority to establish and enforce certain binding rules of common coexistence.
This distinction is crucial.
If law becomes synonymous with morality, everything legal becomes automatically moral and everything illegal automatically immoral. History does not support such a simple identification.
Here Habermas becomes relevant. The legitimacy of norms cannot be reduced merely to the existence of a state authority; it is connected to public reasoning, participation, and the possibility that citizens collectively shape the norms by which they are bound. Habermas developed systematic accounts of communicative action, discourse ethics, law, and deliberative democracy.
Thus, instead of the crude choice between “state surveillance” and “anarchy,” another possibility emerges: institutions of law constrained by rights, legitimized through democratic procedures, and complemented by a civil society possessing strong moral autonomy.
7. Artificial Intelligence: The Great Overseer or the Mirror of Action?
The issue acquires a new historical dimension here.
Artificial Intelligence may dramatically increase the capacity of institutions to monitor, predict, and regulate human behavior.
In such a scenario, the future could mean the technological perfection of the first sentence:
“Don’t do it, because I’m watching you.”
But there is another possibility.
AI could be used to help human beings understand the consequences of their choices, examine different perspectives, identify potential harms, and recognize the individuals and social groups affected by their decisions.
The distinction is fundamental:
AI as an instrument of surveillance
or
AI as a technology for strengthening moral understanding.
The first increases external compliance.
The second may strengthen internal judgment.
The contemporary international debate on AI ethics already emphasizes human dignity, human rights, transparency, accountability, non-discrimination, and human oversight. UNESCO’s global Recommendation on the Ethics of Artificial Intelligence, adopted in 2021, explicitly places human dignity and human rights at its core and emphasizes that AI systems should not displace ultimate human responsibility and accountability.
Yet this does not solve the deeper problem.
Technology can help us know more.
It cannot guarantee that we will care more.
And morality is not knowledge alone.
8. From Human Being to Posthuman?
The concept of the “posthuman” can be employed here not as a technological prediction but as a philosophical hypothesis.
The posthuman would not simply be a human being with greater computational power, greater longevity, or technologically enhanced capacities.
The posthuman could instead be understood as a human being who has overcome a deeper limitation of the human condition: the inability to adequately perceive the consequences of one’s actions upon others.
If this is the horizon, then the transition toward the posthuman will not be determined by technological development alone.
It will depend upon whether technology strengthens human autonomy or replaces it.
A human being who obeys an algorithm because “the algorithm knows better” is not necessarily more morally autonomous than a human being who obeys a state apparatus because he fears punishment.
Real progress would mean something different:
giving human beings a greater capacity to understand, rather than merely a greater capacity to obey.
9. Conclusion: From “I See You” to “You See”
The real dilemma of the age of Artificial Intelligence is not whether we will have more rules or fewer rules.
It is what kind of human being our civilization will create.
A human being who behaves correctly because he is being watched?
Or a human being who behaves responsibly because he has developed the capacity to see the other?
The first proposition is:
“Don’t do it, because I’m watching you.”
The second is:
“Don’t do it, because you understand what you are causing the other.”
But perhaps there is a third, deeper formulation:
“Don’t do it, because you see.”
You see the other.
You see the other’s vulnerability.
You see the consequences of your action.
You see that your freedom does not exist in a vacuum, but within a world shared with other human beings.
This may be the real ethical challenge of Artificial Intelligence.
Not to build a machine that will watch us better.
But to create a technology that will help us become human beings who need to be watched less and less.
The greatest technological progress, therefore, may not be the creation of an intelligence superior to our own.
It may be the creation of conditions under which our own intelligence finally acquires the moral maturity to see the Other.
And then the distance between
“Don’t do it, because I’m watching you”
and
“Don’t do it, because you understand what you are causing the other”
will no longer be merely a path toward a technological posthuman.
It will be a path toward a more autonomous human being.