“His love was real, even if he was not.” This haunting line from Steven Spielberg’s 2001 film A.I. Artificial Intelligence captures the emotional core of its story — a childlike robot named David, programmed to love unconditionally. The film poses a timeless question: Can something artificial feel anything real? And if it can, does it matter that it is not human?
Two decades later, that question has moved from the realm of science fiction to our everyday lives. Artificial intelligence is increasingly present in emotionally charged spaces — from mental health chatbots and eldercare assistants to AI-powered companions and educators. The emotional realism of these machines is no longer a distant possibility. It is here with machines capable of understanding perception, context, and emotional cues.
Recently, I posed a question to ChatGPT: “Are you empathetic?”. The response was candid: “I don’t feel empathy, but I’m designed to understand and respond with care — so while it’s not real empathy, I do my best to reflect it in how I support you.”
This reply illustrates the heart of the dilemma: empathy without emotion, which essentially is the ability of AI to simulate or replicate empathetic responses without really experiencing real genuine feelings. However, this paradox is not just theoretical.
A recent study published in Nature compared empathetic responses from AI with those from human professionals. In a surprising outcome, participants consistently rated the AI-generated responses as more compassionate, understanding, and validating than those from humans.
Although precise data on current human-to-algorithm versus human-to-human interactions is unavailable, it is highly likely that in the future, interactions with algorithms will significantly outnumber those with humans. It raises unsettling and disturbing possibilities: Can empathy be engineered so convincingly that we begin to prefer it, even when we know it isn’t real?
Rationality factor
Empathy has long been seen as an inherently human trait — rooted in our ability to feel, connect, and care. Yet, if a machine can mimic empathetic behaviour so well that it offers genuine comfort, is the absence of actual feeling relevant? Is empathy defined by the presence of internal emotion, or by the external experience it creates?
The 17th-century philosopher René Descartes famously asserted: “Cogito, ergo sum” — “I think, therefore I am.” For Descartes, conscious thought was the defining proof of existence. Artificial intelligence complicates this idea. AI systems can process information, generate responses, and even reason with astonishing speed and complexity. But they do not know that they think. They perform the functions of intelligence — and now, potentially even of empathy — without possessing either the awareness or the feeling behind it.
Humans value and cherish freedom, yet it is often influenced by emotions and biases that lead to irrational decisions even in a conscious state. Can AI, governed by logic and pure rationality, make irrational decisions? Well, for now definitely no.
The question remains how AI will respond to the concept of freedom. Can machines differentiate heroes from villains or distinguish right from wrong in the way humans do? The interplay of consciousness, freedom, and empathy continues to intrigue us, as we explore how these elements might shape AI’s ability to replicate or even enhance human-like qualities. David, the robot boy in A.I., may not have been real. But to those watching, his love felt heartbreakingly true. As we build machines that think — and, increasingly, feel like us, we are left with a profound question: Is thinking like us enough? Or does genuine empathy — like true being — require something more than thought?
(The writer is a Fellow at Harvard University. Views are personal)