Heidegger and Phenomenological Approaches to Work

This week I am sharing a work-in-progress: the draft introduction to a scholarly essay I'm writing for the Oxford Handbook of Philosophy and Work. I am delighted to be included in this distinguished collection of essays, which should appear sometime in 2026. I was invited to contribute a chapter on "Heidegger and Phenomenological Approaches to Work." Please do send me a message if you have any suggestions for improvement.

1. Introduction

Work is central to Heidegger’s Being and Time, the foundational text of existential phenomenology. For Heidegger, work reveals the underexplored and phenomenologically fertile domain of pre-reflective skillful action, the domain of everyday “dealings” [Umgang]. Such activity, which Hubert Dreyfus called “skillful coping,” falls through the cracks of modern philosophy’s ontological divide between subjects and objects.[1] 

Heidegger musters his descriptive phenomenology of work in his project of overcoming modern ontology’s prioritization of an epistemological, knowing subject and the corresponding dichotomies between mind/body, self/other, and self/world. His exemplar is a carpenter at work in his workshop. The carpenter’s familiarity with and ability to get around in his workshop call forth novel ontological categories beyond Descartes’ res cogitans and res extensa. According to Heidegger’s picture, mind and world are part and parcel of an intricate relational structure he calls “being-in-the-world.”[2]    

Even leaving aside Heidegger’s fundamental ontological pretensions, which he himself subsequently abandoned, his account of the significance of work remains timely and important. The nature of human work is today being radically brought into question again due to recent awe-inspiring advances in AI technology, especially large language models (LLMs). The fact that these models have neither an endogenous (or even robotic) embodiment in the world nor an ability to track the truth of their conversational output has not stopped a surge of startling predictions.

Many are saying that AI will soon, finally, after decades of false promises and bad starts, be able to take over most work from human beings. The recent statements of one popular commentator, Ezra Klein of the New York Times, are representative of the current predominant cultural attitude towards AI and work. Klein avows that “we’re about to get artificial general intelligence [AGI] … within two or three years.”[3] 

But what does he mean by AGI here? It is “transformational artificial intelligence capable of doing basically anything a human being could do behind a computer—but better.” If we are to believe such prognostications, the field of AI is finally ready to make good on a variant of a prediction made 60 years ago by one of its major early pioneers, Herbert Simon, who then claimed that “Machines will be capable within twenty years of doing any work that a man can do.”^[4]

The contemporary version of Simon’s prediction includes a crucial qualification. AI is supposed to take over only the kind of labor done by remote workers during the COVID-19 pandemic, the contemporary incarnation of those Peter Drucker designated as “knowledge workers” back in 1967.[5] Thus, in conversations about the fate of work in the face of today’s AI, it is common find appeals to a notion of specifically “cognitive” work. The implied contrast is to embodied, “manual” work (as it was also for Drucker).

In these post-COVID-19 days, however, the contrast class also includes the modes of work specifically designated as “essential” during the pandemic, e.g., caring professions such as childcare, elder care, teaching, nursing, etc., where embodied co-presence and a caring synthesis of felt intuition with abstract reasoning are indispensable for quality performance. Though LLMs are becoming increasingly skilled at writing computer code, nobody is at all tempted to turn over care of their children to AI agents.

Characterizing anticipated AGI systems as being limited to “cognitive” tasks is a reasonable hedge. Moreover, it is a hedge that harkens back to the skepticism Hubert Dreyfus brought to Simon’s predictions in the '60s and '70s.^[6] The main ground of Dreyfus’s skepticism toward Simon’s claims is the phenomenology of embodied, skillful human activity presented by Martin Heidegger in Being and Time. Dreyfus’s critique was aimed at the rules-based, symbolic AI that previously dominated the field, not the now-regnant neural network approach at the heart of LLMs.

However, despite the fact that some of Dreyfus's specific criticisms fail to apply to LLMs, the Heideggerian view of embodied, practical intelligence and agency motivating his critique, and further developed by the tradition he inspired, remains both a powerful contrast to the brittle, disembodied, and uncaring form of intelligence exhibited by LLMs, and an enduring reminder of what skillful, meaningful human work involves: embodied presence, tacit know-how, and committed co-involvement.

 Indeed, in Dreyfus’s hands Heidegger’s phenomenology becomes the foundation of an entire incipient tradition focused on rethinking the nature of work. This chapter explores two key developments of this tradition: the accounts of (1) skill acquisition and embodied know-how as developed by Stuart Dreyfus with Hubert, and (2) the structure of the conversational coordination of human work, as developed by Fernando Flores with Terry Winograd.[7]

I begin by reconstructing Heidegger’s ontology of work in Being and Time. I focus on two of its aspects: pre-reflective, skillful engagement with “ready-to-hand” equipment, and the intrinsically social and conversational nature of human involvement in the world. The subsequent sections examine how Dreyfus/Dreyfus and Winograd/Flores expand upon these themes in their respective accounts of human skill and work.

In our age of AI, we may soon learn that what machines still cannot do is precisely what makes human work meaningful. Human intelligence is not a free-floating disembodied capacity to solve narrowly defined cognitive problems. Our form of intelligence and our capacity for meaningful work are intimately connected to our shared, embodied, skillful, caring, and conversational being-in-the-world.

NOTES

[1] See Hubert L. Dreyfus, Being-in-the-World: A Commentary on Heidegger’s Being and Time, Division I (MIT Press, 1991), and Skillful Coping: Essays on the Phenomenology of Everyday Perception and Action, edited by Mark Wrathall (Oxford University Press, 2014).

[2] My focus here is strictly on Heidegger’s conception of work from the period of Being and Time (BT) (1927). For an interpretation of Heidegger’s post-BT ontology of work, see Blok, “Heidegger’s Ontology of Work” (Heidegger Studies, vol. 31, 2015, pp. 109–128) and Blok, Ernst Jünger’s Philosophy of Technology: Heidegger and the Poetics of the Anthropocene (Routledge, 2017). Blok connects Heidegger’s thinking about work in the 1930s to his notorious and abhorrent involvement with the Nazi regime in Germany in 1933.

Heidegger’s ontology of work in the first part of BT does not exude any particular political commitments, but whether or not there are connections between Heidegger’s philosophy and his politics is of course a huge and hugely important question. For useful discussions, see Iain Thomson, “Heidegger’s Nazism in the Light of his early Black Notebooks: A View from America” in Alfred Denker and Holger Zaborowski, eds, Zur Hermeneutik der ‘Schwarzen Hefte’: Heidegger Jahrbuch 11 (Karl Alber, 2017, pp. 184–209). See also, Thomson, Heidegger on Ontotheology: Technology and the Politics of Education (Cambridge University Press, 2005).  

[3] Ezra Klein, “The Government Knows AGI is Coming,” March 4, 2025, The New York Times. 

^[4] Herbert Simon, The Shape of Automation for Men and Management (Harper & Row, 1965).

[5] Peter Drucker, The Effective Executive (Harper & Row, 1967).

^[6] Dreyfus, “Alchemy and Artificial Intelligence” (RAND Corporation report, 1965) and What Computers Still Can’t Do(MIT Press, 1992 [1972]). Dreyfus’s critique of AI also drew heavily upon the later Wittgenstein, as well as Kierkegaard and Merleau-Ponty. Interestingly, it was Peter Drucker himself, the originator of the notion of “knowledge work,” who recommended Dreyfus’s original 1965 report be expanded and published into a book. See Dreyfus and Dreyfus, Mind Over Machine (The Free Press, 1988, p.71).    

[7] Terry Winograd and Fernando Flores, Understanding Computers and Cognition: A New Foundation for Design (Ablex Publishing, 1986). Owing to space constraints, I can regrettably only gesture at a third crucial contributor to the Dreyfus-inspired, post-Heideggerian tradition devoted to rethinking the nature of work, namely, the account of the skills and education of nurses developed by Patricia Benner. See for example, Benner,From Novice to Expert: Excellence and Power in Clinical Nursing Practice(Addison-Wesley, 1984). 

Hubert Dreyfus once remarked: “Patricia Benner and Fernando Flores are the people that I have worked with who have done more than anybody to teach me that Heidegger has immense applied implications that I had no idea flowed from his ontology” (Introductory Remarks to Fernando Flores’s lecture at the Applied Heidegger Conference, UC Berkeley in September 1989).

What suggestions do you have for improvements of this draft? What questions, thoughts, observations, or perplexities does all of this bring up for you? Let me know in the comments or by sending me a message!

Join the Conversation

Your support makes Without Why possible. If you upgrade to a paid subscription for as low as $8 per month you gain the opportunity to post and reply to comments on these pages, and to join a growing community of conversation here. You will also receive extra-special members-only posts. Sign up today!