# Eindhoven Research Proposal (distilled) This proposal develops an account of how generative AI systems can contribute to human flourishing in creative and intellectual work. On perfectionist accounts of well-being, human flourishing is a matter of exercising and developing one's capacities (Bradford 2015), including in digital contexts (Dennis 2020). Generative AI puts pressure on such flourishing. If Midjourney can produce a finished illustration from a prompt, why train as an illustrator? As LLMs make progress in mathematics, why train as a mathematician? These systems threaten creative flourishing by making certain forms of creation _frictionless_. An image or text can be produced in moments, with little knowledge or effort. Danaher and Nyholm (2021) describe such cases as opening an _achievement gap_, in which the product of an activity is still produced while the person no longer carries out the part of it that was worth doing. On Bradford's (2015) account, difficulty and competence are sources of an achievement's value, so a result supplied by a prompt is not yet an achievement of the person who receives it. The threat is to the user's agency: what do they still do, when the product can be had without them? I shall call the agency at issue — the one a user exercises in drawing results from a partly autonomous system — _elicitive agency_. Characterising it builds on my own work on the phenomenology of agency (Young 2022). Generative AI systems differ from ordinary tools, even digital ones, in two respects. First, control over them is only ever partial: a user has complete control over a hammer or word processor, but an LLM or image generator runs partly of its own accord. Terrone and I have compared such systems to nature, in that they "autonomously generate and dynamically shape things in ways that are so complex and chaotic to appear unpredictable" (Young and Terrone 2025). Second, trained on large bodies of text and images, they are _potent_: able to produce an extraordinarily wide range of outputs. These features mean that eliciting worthwhile outputs is itself a skill: results have to be coaxed from the process, an activity Terrone and I have likened to a gardener growing a garden (ibid.). The skill lies in recognising which of the system's returns answer to an aim it cannot supply, and reworking them toward that aim, rather than accepting the first result offered. Because this skill grows with practice, such work develops the user's capacities as well as exercising them, and so can contribute to flourishing across a life rather than diminish it (Dennis 2020). This recognition is exercised across a working relation that develops over time. The unit of assessment is therefore this relation, and the question is whether it keeps the user at work within it or reduces them to a recipient of finished results. The user's exercise of capacity depends on a partly autonomous system, and the system's worth depends on the practice it sustains. What goes well or badly is this dependence as a whole, involving both the person and the system. When it goes well, a sociotechnical system is flourishing. Two questions follow. How should elicitive agency be characterised? And how should generative systems, and the practices around them, be built and assessed so that they support the exercise and development of a person's capacities rather than only delivering products? Whether a user can work in this way depends on how a system is designed and on the setting into which it is introduced. A system can be built and deployed so that reworking its outputs is natural, or so that accepting the first finished output is easiest. The second question cannot be answered by philosophical argument alone. It is about an actual practice, and those who carry it out understand things about that practice that an argument from outside would miss. The project would therefore be pursued with those who use these systems and those who introduce them into working settings, through co-creation under Bombaerts and within the STEM line, while treating co-creation itself as a practice whose assumptions need scrutiny (Tsui et al. 2025). The conceptual contribution would be an account of the agency a user exercises in working with a partly autonomous system. The practical contribution would be a framework for telling when a generative-AI practice sustains a person’s exercise of their capacities and when it stands in for them. ## References Bradford, G. (2015). _Achievement_. Oxford University Press. Danaher, J., & Nyholm, S. (2021). Automation, work and the achievement gap. _AI and Ethics, 1_(3), 227–237. https://doi.org/10.1007/s43681-020-00028-x Dennis, M. J. (2020). Cultivating digital well-being and the rise of self-care apps. In C. Burr & L. Floridi (Eds.), _The ethics of digital well-being: A multidisciplinary approach_. Springer. Tsui, S., Pfotenhauer, S., & Bombaerts, G. (2025). Constructing the “cocreative citizen”: Innovation subjects, statehood and the common good in hybrid public–private settings. _Science, Technology, & Human Values_. https://doi.org/10.1177/01622439251394488 Young, N. (2022). Agents of change: Temporal flow and feeling oneself act. _Philosophical Studies, 179_, 2619–2637. https://doi.org/10.1007/s11098-022-01790-w Young, N., & Terrone, E. (2025). Growing the image: Generative AI and the medium of gardening. _The Philosophical Quarterly, 75_(1), 310–319. https://doi.org/10.1093/pq/pqae120 # Draft of Eindhoven Cover Letter Dear Dr Dennis and Professor Bombaerts, I am writing to apply for the ESDiT fellowship on sociotechnical system flourishing. The project asks how flourishing can be understood at the level of sociotechnical systems, rather than only at the level of the individual who uses a given technology. My own work has been about the value of using technology — what artefacts let us do, and the forms of agency they make possible — and it has been moving from the individual user toward the practices that technologies make possible. This post asks the same kind of question at the level of the system in which those practices sit. The research plan I am submitting takes up that question through generative AI. These systems produce results that would once have required skill, while asking little visible effort of the user. If flourishing consists in exercising and developing our capacities, then the value of working with such a system cannot be read off the quality of what it produces. The question is whether the practice still leaves the user something worth doing: whether it sustains the user's creativity or substitutes for it. In "Agents of Change" I argued that the sense of oneself as the source of one's action is a feature of agentive phenomenology, tied to temporal flow. In recent joint work with Enrico Terrone, including "Growing the Image" (Philosophical Quarterly, 2025), I have examined artefacts through the activities they support and the agency they make possible. I also teach this material at Genova, in a philosophy of technology course that now includes the philosophy of generative AI. That is my starting point for the project: an account of how artefacts matter through the forms of activity they make possible. Whether a generative-AI practice sustains the user's creativity or replaces it is not something philosophical argument can settle on its own. The answer turns on the practice that builds up between a person and a system over repeated use, since that is where the relevant capacities are exercised and developed, or left to atrophy. That is why the co-creation dimension of the post bears directly on the research question. An account of when such a practice goes well has to be tested with the people engaged in it, because they know where creativity is being exercised and where the system has begun to stand in for it. This is the part of the project I would most want to develop under your joint supervision and within the STEM line. My publications, teaching, and referees are in the CV. My current position at Genova runs until February 2027, though an earlier release may be possible. Yours sincerely, Nick Young