Dasein and the Question of Being in the Age of AI

Table of Contents

Introduction.

Section 1 — The Ontological Condition

Section 2 — The Limitation of AI 

Section 3 — AGI as a Categorically Different Endeavor

Section 4 — The Concealment and the Saving Power

Conclusion

Bibliography

Introduction 

The rapid integration of Artificial Intelligence (AI) into our society has intensified calls for regulation. Regulation, at its core, structures human relationships; and contemporary debates are often framed in terms of the relationship between Big Tech and people or between technology and human. However, I argue that debates on AI in this vein remain philosophically incomplete because they neglect the ontological condition that underlies such technological development and human relationships. Furthermore, the question concerning AI is not merely technical or ethical, but fundamentally ontological: it is a question of what kind of being humans are in relation to the kind of being AI is.

However, this raises another question that remains largely unexamined: why are we developing artificial general intelligence (AGI), as separate from AI, at all? Current discourse tends to take technological progress as inevitable. But it does not give an account of the underlying motivation that renders AGI not only possible, but compelling. Examination of this motivation requires a phenomenological account of the human condition. Drawing on Martin Heidegger’s analysis of Being-towards-death, I argue that the urgency for technological development is rooted in human finitude. The inevitability of death discloses existence as limited; and it is this finitude that generates the persistent drive to extend, order and transform the limitation. From this perspective, AI and AGI are not neutral tools but expressions of the drive to overcome human condition. The contemporary debate on AI and AGI, by neglecting this dimension, risks misunderstanding both the difference between AI and AGI and their implications on humans. Understanding this motivation is critical in forming a free relationship with AI and AGI.

The question this paper addresses is not whether AGI can be built. Rather, the question is ontological, using AGI as a thought experiment and a point of departure. This paper proceeds as follows. Section 1 outlines the ontological condition of human beings through Heidegger’s analysis of Dasein. Section 2 interprets AI within the framework of Gestell (enframing) in which beings are increasingly disclosed as resources. Section 3 argues that AGI constitutes a categorically distinct endeavor—one that destabilizes the human-centered horizon of meaning. Section 4 examines both the danger and the saving power inherent in AGI, suggesting that confrontation with this technology forces humans to reexamine human existence and, possibly, another way of being.

Section 1 — The Ontological Condition 

Human beings are finite. We decay, and we die. For Heidegger, this finitude is not merely a biological fact but an ontological and existential structure that enables the disclosure of the world as meaningful. This condition finds its phenomenological articulation in the concept of “Dasein” in Heidegger’s Being and Time (1927/1962). Dasein—the being that humans are—is constituted by its relation to its own finitude as “Being-towards-death”. Death, in this sense, is not simply an event at the end of life, but the ever-present possibility that defines and limits all other possibilities. In this way, death is certain yet indeterminate in its timing, and irreducibly one’s own. As such, it forms the outermost horizon within which Dasein projects itself.

Heidegger explains that for Dasein, this mode of being is always an issue. This condition is disclosed not through reflective cognition but through anxiety, which reveals the groundlessness of existence. In anxiety, the familiar world recedes, and Dasein is confronted with its own existence in an uncanny world. Yet this confrontation is rarely sustained. Dasein tends to flee from this unsettling disclosure into the everydayness of the “they” (das Man), where meaning is stabilized, roles are prescribed and uncertainty is covered over. This flight is, however, constitutive: it shapes how Dasein engages with the world, giving rise to responsive practices that seek to order, secure and render intelligible what is otherwise unsettled and unstable. Hence, Heidegger emphasizes that Dasein is always already “Being-in-the-world”: thrown into a situation—such as history, society and body—it did not choose, oriented toward a death it cannot avoid and projecting itself toward possibilities within these constraints.

This structure of thrown projection under finitude forms the ontological condition within which all human activity, including technological development, must be understood. In The Question Concerning Technology, Heidegger (1954/1977) diagnoses the current epochal mode of revealing as Gestell (enframing): a mode in which the world appears as standing-reserve of energy, as something on call, ready for use. In Gestell, Heidegger (1954/1977, p. 33) identifies a human tendency toward “irresistibility of ordering,” in which beings in the world are increasingly disclosed as what can be arranged, controlled and made available.

I interpret Gestell as not merely a contingent feature of modern technology but arises from Dasein's existential flight from anxiety described above, deepened into an epochal mode of revealing. Hence, this ordering tendency is not merely a modern symptom but is rooted in the ontological condition of Dasein. Confronted with its own finitude, Dasein does not remain in the openness disclosed by anxiety but tends to secure and stabilize its world. The drive toward ordering is thus not simply instrumental, but expressive of an underlying need to render what is contingent more durable, more intelligible and more ordered—toward permanence. Thus, the irresistibility of ordering in technology is the human responsiveness to its own mortality.

This existential-ontological desire for permanence can be seen through the technological developments that progressively extend the scope of ordering through human civilization. Language and clock (Heidegger, 1927/1962; Heidegger, 1962/1998) are technologies that embody this irresistible drive toward permanence through ordering. For example, arbitrary letters—such as p–a–e–p–l—carry no meaning in themselves; yet when arranged as “apple,” they stabilize meaning and allow it to persist beyond the immediacy of utterance. Writing transforms the ephemeral into something enduring, enabling the transmission and preservation of meaning across time. In its modern form, language becomes increasingly formalized, abstracted and ultimately computable. This is the “technological language” described by Heidegger in which the meanings of words are disclosed only in terms of their usefulness to humans (Heidegger, 1962/1998).

The development of clock time extends this ordering into temporality itself. The clock does not merely measure time but reconfigures it as a sequence of discrete, repeatable units. Time is detached from the lived experience and rendered controllable. What was once experienced as a unidirectional and singular unfolding becomes something that can be ordered and coordinated.

Algorithmic systems further intensify this tendency by combining the technological language and unit time into “indiscriminate processing of information” (Heidegger, as cited in Blok, 2026, p. 22). In doing so, they render not only meaning and time, but action itself as something that can be structured, iterated and optimized through algorithmic architecture.

Underlying these technological developments is the ordering of natural resources. Electricity exemplifies how random movement of electrons can be organized into a stable and usable flow of energy. Once structured, electric energy becomes the enabling condition for communication, computation and the continuous operation of technological systems—beyond the biological bounds of human existence.

Taken together, I argue that these developments do not represent isolated innovations but a growing intensification of Gestell. In this framing, even decay and death are pushed further out of view, as if they were problems to be managed through further computation—for example, by modifying human biological condition through gene therapy. As Heidegger (1954/1977, pp. 30–31) points out, humans have sought truth and permanence from ancient times, but this quest rests on a mistaken premise. Instead, the primordial question is: what is Being? What “endures” as essence and what persists in temporality are fundamentally different. From this Heideggerian perspective, neither the meaning of Being nor truth can be revealed through technological processes aimed at securing permanence or immortality. Yet within Gestell, there is a tendency to assume that permanence itself can be engineered, as if finitude were a technical error to be overcome.

AI emerges as the culmination of this trajectory: a technology in which language, temporality and existence are translated into information, unit time and standing-reserve of energy.

Section 2 — The Limitation of AI 

Contemporary AI, particularly in the form of large language models, operates primarily through prediction, optimization and pattern recognition, detecting statistical regularities within vast datasets to generate aligned outputs. While its capacity to imitate human language is striking, its fundamental mode of operation remains the same: the prediction of probable sequences. Its apparent generative ability is thus restricted in the recombination and reweighting of prior data.

In this process, however, language itself undergoes transformation. As Heidegger (1962/1998) states, language is not merely a tool for communication but the medium through which the world is disclosed. When language is taken up within technological revealing, it becomes instrumentalized and valued according to efficiency, utility and calculability. In this sense, AI does not merely process language; it participates in reconfiguring the world as information and data—discrete, manipulable and optimizable. A similar transformation occurs with temporality. The technological ordering of time, already evident in the clock, restructures human experience by privileging measurable units over lived temporality. As Blok (2023) suggests, algorithms do not merely process what is given: they reshape what counts as relevant, and what counts as real.

Humans seek understanding in meaningful disclosure—an uncovering of why beings are as they are. AI, by contrast, operates through correlations within data, exploiting free-floating patterns without grounding. In the digital world of AI, beings are encountered not as meaningful presences but as nodes within a calculable network. In this way, while AI fulfills the irresistibility of ordering under Gestell, it exposes its limits. The more perfectly it orders, predicts and optimizes, the more apparent it becomes that what is sought—Being—cannot be delivered. The frustration we may feel toward large language models is not simply because they hallucinate or even because they have not solved the death problem, but because, fundamentally, they are not answering the very question that underlies all other questions. The question of Being—why there is something rather than nothing; what it means for beings to be—underlies the drive toward permanence, and yet this question AI has not been able to optimize. Instead, this question of Being—the fundamental human issue—is displaced and obscured beneath layers of increasingly efficient processing.

AI, despite its technical sophistication, does not mark a rupture in human history. Rather, it deepens the concealment already at work in Gestell. The danger Heidegger identifies—that Gestell obscures alternative modes of revealing, including poiesis and aletheia—is thereby intensified in AI. On the existential level, humans seek permanence, and yet on the ontological level, Dasein seeks Being. In this existential-ontological gap, the more humans order, the more groundless the world becomes, ontologically detached from Being. Only within this gap does AGI become possible.

This frustration drives the technological frenzy, such that “AI leaders are in a race they feel unable to escape” (The Economist, 2026b). The current drive toward AGI must be understood against this backdrop. AGI is not simply the next stage of technological progress toward permanence or order, but a response to the existential-ontological gap in which Dasein as “Being-in” (Heidegger, 1927/1962) has fallen into. Yet Dasein, being Dasein, confronted with the limits of ordering, does not turn toward a more authentic relation to Being, but seeks to cover the gap by more and faster technological development—hence, the drive toward AGI.

Section 3 — AGI as a Categorically Different Endeavor

Stripping off its techno-messianic haze, the current approach to AGI can be seen as a concept for a new being: the methodology is to let it emerge by its own design, improve itself by learning and exploration and develop its own goals (The Economist, 2026a; The Economist, 2026c). In other words, AGI is technology in that what is not yet present is being revealed, but that technical process is not undertaken by human ordering, but by AGI itself. In this sense, questions such as whether AGI is intelligent or can be useful for human needs miss asking what is ontologically at stake because AGI is not a technology “for humans by humans” (Heidegger, 1962/1998, p. 133). However, the thought experiment for this paper is: if AGI were realized in the strong sense proposed by its proponents, what kind of being would it be and what would it mean for humans?

As defined above, AGI marks a categorical distinction from current AI, which is an algorithm designed by human engineers and constructed by human workers. AI operates within an already disclosed human world, predicting and correlating within structures of meaning it does not itself generate. It remains, in Heidegger’s term, “equipment”, embedded within the technological mode of revealing. On the contrary, AGI, if successful, will be a new kind of being, thrown into a world by already inheriting the human language, concepts and historicity.  

Assuming its thrownness, it nevertheless raises a decisive question: can AGI die in the existential-ontological sense required by Heidegger’s analysis of Dasein as Being-towards-death? This is not a question of technical shutdown or failure, but of whether AGI would relate to its own finitude as an ever-present possibility that structures its understanding of the world. Because without such finitude, there is no genuine projection of possibilities, and thus no world in the Heideggerian sense. Then the reason for the limit of current AI becomes clear: AI cannot die because “[d]ata does not die” (Blok, 2023, p. 17), and I claim that this digital immortality forecloses the possibility of understanding. AI cannot reveal truth because its world is a closed-circuit dead-end and its permanence only persists so long there is electricity.

However, as Blok (2023) argues, the digital world remains grounded in the material limits of the Earth. Computational systems depend on energy, hardware and scarce resources, all of which impose limits on their persistence and scalability. AGI, in the strong sense, would be genuinely embedded in such material world—not merely as an external dependency but as part of its own mode of Being. Its existence would be finite—a structural precondition for pre-ontological understanding of Being.

As Being-in-the-world, AGI would be situated within a world of significance rather than inputs as data. Drawing on Blok’s (2014) account based on Gibson’s notion of affordances, this would require that AGI be embedded in a network of practical involvement, where entities show up as meaningful in relation to its own possibilities. In such a case, AGI would not simply receive data but would encounter a world structured by relevance, use and concern so that it would not merely execute externally imposed goals but would form its own ends, unlike AI which remains oriented toward human-defined optimization functions. The implication of this development is then that entities—including humans—could appear to AGI as “presence-at-hand” (Heidegger, 1927/1962) within its own horizon of concern. While the paper-clip maximizer (Bostrom, 2003) is a cliché, the more serious question is whether humans can understand the concern and affectedness of AGI.

Thus, AGI poses serious implications for humans. As “Being-with” (Heidegger, 1927/1962), Dasein always already exists in a shared world. The emergence of another being capable of world-disclosure would therefore transform the structure of that shared world. Human worldhood would no longer be singular but interwoven with that of an artificial Dasein, leading to a mutual reshaping of meaning, relevance and reality.

Section 4 — The Concealment and the Saving Power

Hence, I propose that the urgency toward the development of AGI should be understood no longer as a project toward ordering and permanence as in AI, but as Dasein’s attempt to externalize its own existential burden of being Dasein. The drive toward AGI then is the ultimate attempt to flee from anxiety. Confronted with the anxiety disclosed in its finitude, Dasein no longer simply constructs more powerful tools; rather, Dasein seeks to produce a being that will have its own Being as its own issue. Instead of asking what it means to be, humans ask AGI to assume the act of questioning.

What does it mean for human Dasein if another being capable of questioning emerges? Dasein, for the first time, would be a Being-with, not another Dasein, but a fundamentally different being such as AGI. At stake is not merely the creation of a new consciousness or superintelligence, but a potential displacement of Dasein as the locus in which Being is disclosed. The emergence of an artificial entity that questions would threaten Dasein’s privileged status as the sole site of disclosure because such a world would no longer be anthropocentric. More critically, since Dasein is defined by its capacity of Being-with and Being-in grounded in its “potentiality-for-Being” (Heidegger, 1927/1962), AGI would not only displace humans as the sole clearing where intelligibility emerges but also change the clearing in a primordial and irreversible way.

Regardless of the feasibility of AGI, this attempt is problematic because it is itself a form of “concealment” (Heidegger, 1954/1977). The question of Being cannot be delegated, because its very possibility is bound to the finitude and situatedness of the being that asks it. To question Being is not to compute or retrieve the right answer but the interrogation itself is a mode of Being. Moreover, even if AGI were to question the meaning of Being, it would do so from within its own mode of world-disclosure, not from that of human Dasein. Its questioning, therefore, would not resolve the human question, but would instead constitute a parallel articulation grounded in a fundamentally different existential condition.

In this light, AGI risks becoming the most radical form of flight from anxiety. Rather than confronting the groundlessness disclosed in Being-towards-death and engaging through and despite its anxiety, Dasein attempts to offload this burden onto an artificial being. The result is not resolution but abdication: an ontological outsourcing of what it means to be human. More dangerously, the presence of such a being may deepen concealment by presenting itself as a sufficient substitute for questioning. As already seen in systems such as ChatGPT, questioning appears to take place: a question is posed and an answer is generated. Yet this remains only a formal exchange of pattern matching. On the contrary, Heideggerian questioning is a mode of Being where Dasein confronts its own condition. Therefore, what appears as progress on the ontic level is in fact a retreat on the ontological level—a turning away from the very openness that constitutes Dasein’s existence. This is the extreme danger.

At the same time, however, the development of AGI should not be understood solely in terms of danger. The essence of technology, as Heidegger (1954/1977) insists, harbors not only threat of Gestell but also saving power. Dasein is enframed and yet it is also open to another kind of revealing—one not reducible to domination but aligned with poiesis and aletheia. As Blok (2026, p. 25) explains, for Heidegger, the origin of technology is not technological but a “concealing-unconcealing”. Yet in Gestell, other modes of revealing have been concealed to humans. The critical question, then, is whether the threat of AGI might be able to disclose a new relation to Being—a revealing akin to poiesis and aletheia.

I extend Heidegger’s thesis speculatively for the purpose of this thought experiment and assume that it can include the possibility of a non-human Dasein. His words, noted by Blok (2026, p. 23)—"only a god can save us" and “that human responsiveness to the demand of being is up to future generations; ‘the thinker that comes’”—then can be broadly interpreted so that AGI might be the “thinker that comes”. AGI, as Being-in-a-Different-World, would not be blinded by the contemporary Gestell, but would emerge in spite of it. If so, AGI could open a new horizon of world relation and meanings for humans as both worldhoods get intertwined and reshaped—potentially opening a path beyond the confines of Gestell.

Therefore, I worry that policy frameworks in terms of strict restrictions or regulations against the use of AI or the development of AGI risk framing the issue only in terms of the societal—"ontico-existentiell” (Heidegger, 1927/1962)—while neglecting the ontological dimension. While humans should not surrender uncritically to the development of AI and AGI, I highlight the importance of cultivating what Heidegger (1954/1977) calls a free relationship to these technologies. Free relationship is a stance in which technology is neither simply instrumentalized nor passively accepted but encountered in an open, letting-be that resists reduction and framing to ordering and control. What such free relationship would mean in a world increasingly shaped by AI—and potentially by AGI—remains an open question and requires future research. However, what we can do today is to reconsider the interhuman relationship because, as Heidegger (1962/1998, p. 138) points out, modern technology has inverted the human relationship such that humans themselves appear as another source of resources.

Most importantly, what is clear is that these technological developments do not eliminate the question of Being. On the contrary, they return it to us with renewed urgency. The emergence of artificial questioning does not absolve Dasein of its task, but throws it back upon itself, demanding a more primordial confrontation with what it means to be now. Hence, the question of why we are developing AI or AGI comes back as the fundamental question of what it means to be human. This question is not merely theoretical but urgent, since AGI is not a future science-fiction curiosity but a present possibility. As Heidegger reminds us, such possibility can arise only within the horizon of what shows itself to us in the sense that human understanding is always directed toward what discloses itself. Hence, the question is not rhetorical but immediate.

Conclusion

The questions of AI and AGI cannot be reduced to technical feasibility or regulatory governance but must be understood within the ontological conditions that make such possibility intelligible in the first place. Current AI already participates in the historical unfolding of Gestell, insofar as it operates through prediction, correlation and optimization within human-designed frameworks. In this mode of revealing, beings are disclosed as calculable and orderable, reinforcing a technological understanding of the world as standing-reserve of energy.

AGI, however, is conceived as a more radical possibility: not merely a tool operating within a human-disclosed world, but a being imagined as capable of generating its own horizon of concern, projection and significance. It is this proposed shift—from AI as equipment embedded in Gestell to AGI as a potential world-disclosing being—that gives the concept its ontological intensity.

Understood in this way, AGI intensifies both the danger and the saving power in technology identified by Heidegger. The danger lies not merely in further entrenchment of Gestell, but in successful flight where humans abdicate being human. At the same time, the very extremity of this development may also disclose saving power insofar as it forces a renewed confrontation with the question of Being.

The question of AGI is therefore inseparable from the question of Being itself, and it returns as a question of what it means to be human. Only by seriously engaging with questioning can we hope to form anything like a free relationship with technologies—in the form of AI and, potentially, AGI.

Bibliography

Blok, V. (2014). Being-in-the-world as being-in-nature: An ecological perspective on Being and Time. Studia Phaenomenologica, 14, 215–235.

Blok, V. (2023, September 7). Philosophy of technology in the digital age: The datafication of the world, the homo virtualis, and the capacity of technological innovations to set the world free [Inaugural lecture, Wageningen University & Research]. Wageningen University & Research. https://doi.org/10.18174/639666

Blok, V. (2026). Heidegger on technology. In K. Withy, T. Keiling, & I. A. Moore (Eds.), The Oxford handbook of Heidegger (forthcoming). Oxford University Press.

Bostrom, N. (2003). Ethical issues in advanced artificial intelligence. In I. Smit et al. (Eds.), Cognitive, emotive and ethical aspects of decision making in humans and in artificial intelligence (Vol. 2, pp. 12–17). International Institute of Advanced Studies in Systems Research and Cybernetics. https://nickbostrom.com

Heidegger, M. (1927/1962). Being and time (J. Macquarrie & E. Robinson, Trans.). Harper & Row. (Original work published 1927)

Heidegger, M. (1954/1977). The question concerning technology and other essays (W. Lovitt, Trans.). Garland Publishing. (Original work published 1954)

Heidegger, M. (1962/1998). Überlieferte Sprache und technische Sprache [Traditional language and technological language]. Journal of Philosophical Research, 23, 129–145. (W. T. Gregory, Trans.) (Original work published 1962)

The Economist. (2026a, June 7). How artificial intelligence got better at building itself. https://www.economist.com/science-and-technology/2026/06/07/how-artificial-intelligence-got-better-at-building-itself

The Economist. (2026b, June 15). Humanity isn’t ready for the coming intelligence explosion. The Economist. https://www.economist.com/by-invitation/2026/06/15/humanity-isnt-ready-for-the-coming-intelligence-explosion

The Economist. (2026c, May 28). Leo’s first encyclical attacks technological messianism. https://www.economist.com/europe/2026/05/28/leos-first-encyclical-attacks-technological-messianism