Sound Research WIKINDX

     p. 60  

AI invites re-thinking of minds (My emphasis)

Clowes, Gärtner, and Hipólito (2021) have argued that we now face a mind-technology problem, as a successor of the mind-body problem: new conceptualizations about the nature of mind and its relationship to artefacts have given rise to ‘a new constellation of basic philosophical problems about the very nature of mind’. For instance, today computers and AI challenge us to reframe and rethink what human minds are. If AI can carry out tasks that previously were seen as belonging to human thought, what does that mean for our understanding of ourselves? Do minds work like computers or not, and why? For example, are minds computational? If not, is it nevertheless a productive metaphor or not? How, exactly, should we conceptualize the relationship between humans and their artificial creations? Are we becoming cyborgs? How are our minds changing as a result of our interactions with the technologies? And what kind of knowledge and self-knowledge does AI create? Does AI distort our epistemic environments? Is knowledge in crisis?

  Added by: alexb44  04/11/2026, 08:32

Comments:

Most of these hardly seem like new questions, but the final three are interesting:

- And what kind of knowledge and self-knowledge does AI create?

- Does AI distort our epistemic environments?

- Is knowledge in crisis?

This seems to very much lean into what Boden talks about, but ultimately to me it still speaks very much to a Stieglerian understanding, and equally to embodied cognition. The technologies we use affect us and vice-versa, but they don't do so in isolation.. they change our view of the world (potentially expand, potentially constrain, imo). 

The question I find most interesting here is how AI distorts our episteme. This is certainly something argued by Stiegler, but he doesn't address exactly how GenAI does this (passed in 2020, so didn't live to see GenAI)

  Added by: alexb44  04/11/2026, 08:32

Keywords:   Agency, Artificial Intelligence, Epistemic Agency, Philosophy of technology

     p. 60  

LLMs are different because of natural language (social epistemology)

"An interesting exception in recent work in epistemology that addresses these issues and cites some works in philosophy of technology is Freiman’s (2023) paper: in order to analyse how conversational AIs may influence people’s beliefs, Freiman proposes the conception of ‘technologybased belief’, which is meant to acknowledge that non-human agency can function as the originator of our beliefs and their content (without assuming that the technology has human-like agency). He argues that technology-based belief is different from testimony-based beliefs, which have a human as the source of knowledge, but also different from instrument-based beliefs, since in the case of conversational AI, the content is delivered in natural language. This is an example of how conceptualizing the influence of technology – here AI – can be integrated with (social) epistemology"

  Added by: alexb44  04/11/2026, 08:36

Comments:

This relates also to the psycho-social-techno-logical view of Stiegler, but not in a way he imagined. Turkle (2026) talks about the myriad ways we are swept into the lure of chatbots from a psychological perspective, and Bender and colleagues (On the Dangers of Stochastic Parrots: Can language models be too big? 🦜) speak about how we envision a mind behind the text.

Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. 2021 On the dangers of stochastic parrots: Can language models be too big? 🦜. Paper presented at Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency.
Turkle, S. (2026). Artificial intimacy: Who we become when we talk to machines. First Edition ed. New York: Little, Brown and Company.   Added by: alexb44  04/11/2026, 08:36

Keywords:   Agency, Artificial Intelligence, Epistemic Agency, Philosophy of technology

     p. 60  

Importance of the technological aspect for epistemology

"Assuming that we wish to be epistemic agents and that epistemic agency is influenced by the social, in this paper, I am interested in analysing and understanding cases in which an agent is, in principle, free to form her beliefs, but in which this formation is rendered more difficult because of technologically shaped conditions. In other words: I’m interested in cases in which epistemic agency is compromised and diminished due to the use of AI.

To ask this question regarding technology is important as discussions in epistemology usually ignore the fact that most of our epistemic lives take place in environments that are very much mediated by digital technologies, including AI and data science. Our acquisitions and constructions of knowledge is not only a social matter but also often takes place via search engines, in social media environments, via the use of electronic devices, and so on. It is therefore advisable to analyse the epistemic consequences of the use of digital technologies for the epistemic work individuals do within social-epistemological environments such as digital social media. Epistemology can learn here from philosophy of technology, media studies, and related fields"

  Added by: alexb44  04/11/2026, 08:54

Keywords:   Agency, Artificial Intelligence, Epistemic Agency, Philosophy of technology

     p. 60  

A wish to be epistemic agents

"In any case, whether we as a matter of fact have full voluntary control over our beliefs or not and regardless of the implications for responsibility, it is clear that as human beings we have the capability to detach ourselves from our beliefs to some extent and to reflect on them, and that most of us tend to value reflective knowledge (Sosa 1991). Most importantly, we like to take an active role in belief acquisition: we wish to be epistemic agents"

  Added by: alexb44  04/11/2026, 08:56

Comments:

LLMs foster the Dunning-Kruger effect more than any other technology - gives the illusion that we have an understanding of things, but this is not real reflective knowledge, it merely looks/feels like it. It is not actually an active role in knowledge acquisition, and we are really surrendering our epistemic agency.   Added by: alexb44  04/11/2026, 08:56

Keywords:   Agency, Artificial Intelligence, Epistemic Agency, Philosophy of technology

     p. 61  

Shifting where knowledge 'exists'

My focus here is on beliefs, but note that the use of AI might also create problems for taking responsibility for our actions – a possibility that also relates to the concerns of epistemology. For example, if we are not able to explain how an AI system (in particular, a deep learning system) formulates its outputs, we suffer from a particular type of ignorance: we don’t know why we are doing what we are doing. This is not only a problem for accepting responsibility for our actions, but also makes it more difficult, if not impossible, to act responsibly towards others (Coeckelbergh 2020)

 

  Added by: alexb44  04/11/2026, 08:25

Comments:

If agency and intelligence are viewed as part of a network, then both agency and intelligence are reticulated in a sense. For any epistemic system, I would argue it is essential that whatever or whoever primarily 'wields' the agency and the intelligence need to be capable of understanding the information. 

Books contain static information only in the sense that the symbols themselves do not change. Yet they change depending on who reads them. LLMs as a source of tertiary retention are always changing - they leave no means for discussion, because they paraphrase everything. They cannot be viewed as a search engine, because any search is accompanied by a description of the relevance of whatever it spits out.

  Added by: alexb44  04/11/2026, 08:25

Keywords:   Agency, Artificial Intelligence, Epistemic Agency, Philosophy of technology