Machines, Morality, and Responsibility: A Dialogue on Ethics in AI

The real question is not whether machines think but whether men do. The mystery which surrounds a thinking machine already surrounds a thinking man.

— B.F. Skinner, Contingencies of Reinforcement; A Theoretical Analysis

Two professors, Ethos and Epistem, sit in a quiet university courtyard, their voices carrying the weight of centuries-old debates reframed for the 21st century. Ethos, the ethicist, leans forward with concern etched across his brow. Epistem, the epistemologist, analytical and precise, adjusts his glasses before responding. A curious student soon joins them, eager to witness the unfolding intellectual duel.

Part I: Problems Caused by AI

Prof. Ethos:
“We cannot ignore the mounting problems AI has unleashed into society. Consider algorithmic bias: AI systems trained on flawed or incomplete data have amplified discrimination. Hiring algorithms once systematically rejected women for technical roles because the data reflected past gender imbalances. Facial recognition software performs poorly on darker-skinned individuals, leading to wrongful arrests (Buolamwini & Gebru, 2018). Bias is not hypothetical , it is tangible harm, Epistem. And we cannot ignore these crimes against human dignity.”

Prof. Epistem:
“Crimes? Nonsense. Machines do not commit crimes; people do. You confuse deterministic systems with malicious intent. The fault lies with engineers and institutions, not the algorithms themselves.”

Prof. Ethos:
“Ah, but that’s exactly my point! And yet you speak as if the cold comfort of determinism absolves us of responsibility. If AI is just math, then why are people dying in Tesla crashes, why are deepfakes destabilizing elections? Math or not, people bleed. Self-driving cars have already caused fatalities, like Uber’s autonomous vehicle striking a pedestrian in 2018 .

Autonomous trading algorithms have triggered market ‘flash crashes’ that wiped billions in minutes . AI-driven misinformation has manipulated elections, from deepfake videos to recommendation algorithms amplifying conspiracies. And even in warfare, autonomous drones threaten to loosen human control over lethal force.”

Prof. Epistem:
“And I say again, Ethos, you exaggerate. One accident, and you want to halt progress? By your logic, we’d have banned cars after the first crash, or electricity after the first electrocution! Technology evolves by error and correction, not by moral panic.”

Prof. Ethos:
“Of what use is the technology if lives are lost instead of betterment. Even if machines are programmed, even if engineers are to be blamed and not machines, that doesn’t change the fact that the mankind is on the verge of destroying itself. Not by creating machines, but by neglecting its faults and instead focusing on commercialisation of AI”

Part II: The Deterministic Nature of AI Agents

Prof. Epistem:
“You don’t listen to anyone, do you? These systems are deterministic in nature. Every AI model, be it a decision tree, a neural network, or a transformer , is bound by mathematical functions. Given the same input and internal state, the output is inevitable. There is no ghost in the machine. Stochastic elements, like random initializations, are themselves governed by probabilistic rules, which are still deterministic in aggregate.”

Prof. Ethos:
“You really are convinced that machines can’t do any harm, aren’t you? But when these probabilities become senses, you would understand”

Prof. Epistem:
“You kidding me!! AI is far from perfect, and the machine consiousness can’t be seen in foreseeble future, or maybe its unreachable. So the belief that AI will one day be ruling us is the most absurd thing I ever heard. Didn’t expect you to fall for it, Ethos.”

Prof. Ethos:
“Then this determinism makes the ethical case stronger, does it not? If AI agents cannot be free, then only humans : the designers, the deployers, bear responsibility for consequences.

It was Prometheus who was punished by the gods for stealing and giving fire to the mankind, not the men themselves. Because they knew that men would destroy the humanity by that fire. And now, that fire is Artificial Intelligence”

Prof. Epistem:
“Indeed. You can’t punish the water for bringing flood, you should punish the men for not building the darn dam.”

Part III: Ethical Responsibility of Humans

Prof. Ethos:
“A dam… Is that all this is to you, Epistem? A mere engineering problem? We are talking about justice, dignity, and the erosion of truth. These are not technical specifications on a blueprint!”

Prof. Epistem:
And what good are your moral proclamations without a blueprint? Justice and dignity are the goals, Ethos, but the mechanism for achieving them is sound engineering and rigorous governance. The solution isn’t to stop building rivers of progress; it’s to mandate better dam engineering. This is why frameworks like the EU’s Ethics AI are so critical — not because they are moral documents, but because they are attempts to create verifiable standards for safety, transparency, and accountability.

Prof. Ethos:
“I agree on the need for standards, but you strip the soul from the argument! These regulations were not born in a sterile lab; they were written in response to real harm. Look at the Cambridge Analytica scandal. The failure there wasn’t just technical; it was a profound moral failure of a company that viewed people not as citizens, but as data points to be manipulated. Your ‘verifiable standards’ will always be five steps behind a corporation’s desire for profit unless they are guided by a strong ethical compass.”

Prof. Epistem: “On that, we are in violent agreement. Corporate self-regulation has proven to be an abject failure. Their internal ethics boards are often just public relations. My point is that the ethical compass is useless without a rudder and a map. We need both: your moral imperative to act, and my insistence on auditable, enforceable systems. Without both, we are lost.”

At this moment, a student, Master Denkerstein, curious and hesitant, approaches.

Master Denkerstein:
“Professors, forgive me for interjecting. But your entire argument for accountability rests on the idea that humans are free moral agents. You must also acknowledge that we ourselves are subject to deterministic forces. Our culture, our upbringing, our very neurobiology. Perhaps our responsibility is not rooted in some metaphysical freedom, but in a practical stewardship: recognizing that while we may be shaped by causes, we still must choose the frameworks that minimize harm.”

Part IV: Deterministic Nature of Reality

Master Denkerstein:
“Consider Libet’s experiment in the 1980s. He measured brain activity before participants reported their conscious decision to act. The readiness potential , a spike of neural activity , appeared milliseconds before subjects felt they had ‘chosen’. This suggests that even our conscious will may be an afterthought, a story our mind tells itself after the brain has already initiated an action.”

Prof. Ethos:
“I am well aware of the experiment,and its challenging implications. But Libet himself argued for a ‘veto power’ — that while consciousness may not initiate decisions, it can inhibit them. Even if determinism holds sway, we exercise responsibility through restraint. The same applies to AI governance: we may not control every causal chain that leads to a new technology, but we can and must veto its harmful deployments.”

Master Denkerstein:
“A compelling analogy.So Determinism at the micro-level need not paralyze moral responsibility at the macro-level. Even if our will is conditioned, we can establish institutions, laws, norms, checks, that function as systemic vetoes.”

Prof. Ethos:
“And that is precisely the bridge. The deterministic nature of AI mirrors the deterministic nature of human thought. But responsibility persists — not because of metaphysical freedom, but because of ethical necessity. We cannot control causality itself, but we can control how we build and regulate its manifestations.”

Part V: Bridging the Gap

Master Denkerstein:
“Then allow me to ask the final question. If both AI and humans are, in their own ways, deterministic, then what grounds do we have for holding anyone responsible at all? Isn’t it all just a chain of cause and effect?”

Prof. Ethos:
“An excellent and vital question. Responsibility is not about escaping causality; it is about creating accountability structures within it. We do not punish a murderer to prove that he had metaphysical free will. We do so to deter future harm, to uphold social order, and to affirm the value of the life that was lost. Responsibility is a pragmatic tool. It is how a society protects its future.”

Prof. Epistem:
“Exactly. Determinism does not absolve us; it contextualizes us. Knowing that an AI is deterministic means we have a duty to design transparency into it: explainable models, rigorous bias audits, and immutable accountability trails. Knowing that humans are influenced by unconscious processes and flawed incentives means we must embed ethics into our institutions and laws rather than relying on the fleeting goodwill of individuals or corporations.”

Prof. Ethos:
“Thus, bridging the gap is ultimately about humility. We must recognize that neither humans nor AI are free-floating agents of pure autonomy. Both are products of complex causes. Yet only humans can reflect on these causes, and from that reflection arises our unique responsibility. We are the custodians of all deterministic systems — biological and artificial alike.”

Student:
“So the lesson is not whether we are free, but whether we are wise enough to steer deterministic systems toward justice?”

Prof. Ethos (smiling):
“Precisely. Wisdom, not freedom, is our ethical horizon.”

Conclusion

AI is neither demon nor savior — it is deterministic computation scaled to societal magnitude. The problems it has already wrought — bias, accidents, misinformation, economic disruption — are not anomalies, but reflections of our negligence. AI agents are deterministic, but so too are we. The challenge, then, is not to deny determinism but to embrace ethical responsibility within it. Through regulation, institutional oversight, and moral reflection, humanity can steer AI’s deterministic power toward human flourishing rather than harm.