Anti-fascist AI

McQuillan, Dan

doi:10.51952/9781529213522.ch007

Cited by 17 publications

(26 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…One possible outcome of the kind of deeper understanding fostered by openness is a call for responsibly limited technology [23,36]. The spectre of regulation (a key way to keep corporate powers in check) is a powerful incentive for companies to keep things proprietary and so shield them from scrutiny.…”

Section: Discussionmentioning

confidence: 99%

Opening up ChatGPT: Tracking openness, transparency, and accountability in instruction-tuned text generators

Liesenfeld¹,

Lopez²,

Dingemanse³

2023

Proceedings of the 5th International Conference on Conversational User Interfaces

View full text Add to dashboard Cite

show abstract

Section: Discussionmentioning

confidence: 99%

Opening up ChatGPT: Tracking openness, transparency, and accountability in instruction-tuned text generators

Liesenfeld¹,

Lopez²,

Dingemanse³

2023

Proceedings of the 5th International Conference on Conversational User Interfaces

View full text Add to dashboard Cite

show abstract

“…Without getting into a full‐blown analysis of the imaginaries and problematisations here, just looking at how Cognii's marketing indicate problem‐solution configurations, such as ‘student success', ‘instructor productivity’ and ‘cost‐effective and adaptive online learning systems and formative student feedback’, clearly shows that not only the discourses (see Blikstein et al, 2022 for an example) need to be studied as problem‐solution configurations but also the actual tools and their sociomaterial underpinnings, operations and effects. This includes the general problem that AI can be trained on data that is biased or flawed, perpetuating discrimination; that it black‐boxes and legitimises certain models of reality and existing relations of power and that it may have significant ecological impacts (McQuillan, 2022). We, therefore, argue that technologies enter education as problem–solution configurations that need to be analytically unpacked.…”

Section: Problematisations In Imaginariesmentioning

confidence: 99%

Imaginaries and problematisations: A heuristic lens in the age of artificial intelligence in education

Rahm

Rahm‐Skågeby

2023

Brit J Educational Tech

View full text Add to dashboard Cite

This paper suggests that artificial intelligence in education (AIEd) can be fruitfully analysed as ‘policies frozen in silicon’. This means that they exist as both materialised and proposed problematisations (problem representations with corresponding solutions). As a theoretical and analytical response, this paper puts forward a heuristic lens that can provide insights into how AI technologies (or advocated AI technologies) function as proposed solutions to certain problematisations based on various imaginaries about how education and learning are best performed or supported. The combined reading of imaginaries and problematisations can thereby aid in our understanding of why and how visions of learning and education are framed in relation to AIEd developments. The overall ambition is to advance theoretical and analytical approaches towards an educational system which is (anticipated as) increasingly permeated by AI systems—systems that also support and implement, more or less, invisible models, standards and assessments of learning, as well as more grand visions of (technology‐augmented) education in society. Practitioner notesWhat is already known about this topic Artificial intelligence in education (AIEd) is repeatedly presented as a solution for a range of educational ‘problems’. This means that such ‘solutions’ must also frame certain aspects as ‘problems’. Such problems and ‘solutions’ (problematisations) also exist within certain imaginaries of the present times and of the future, where these problematisations are presented as particularly significant and acute, and promoting specific anticipations of learning and ideals of education. What this paper adds An exposition of problematisations in educational settings. An exposition of educational imaginaries. A heuristic lens for understanding the ‘present’ and ‘future’ in a particular imaginary as entangled in, and dependent on, a certain ‘past’. Implications for practice and/or policy The approach presented in this paper provides a heuristic lens for examining how AI technologies (or advocated AI technologies) function as proposed solutions to problematisations based on imaginaries about how education and learning are best performed or supported. This aids our understanding of how and why certain visions of learning and education are framed in relation to AIEd developments (real or imagined). It also advances theoretical and analytical approaches towards an educational system, which is (anticipated as) increasingly permeated by AI systems—systems that also support and implement, more or less, invisible models, standards and assessments of learning, as well as more grand visions of (technology‐augmented) education in society.

show abstract

“…Just as modern theorisations of scientific objectivity came hand in hand with new expectations for an ‘objective self’ (Porter, 2014), the technical tendencies of data-driven systems are often wound up with hardline attitudes about data's objectivity that critical researchers have described as ‘enchanted determinism’ (Campolo and Crawford, 2020) or the cathedral of computation (Finn, 2017). At its logical limit lies what Dan McQuillan calls a culture of ultrarationalism in data-driven thinking : ‘a sociopathic commitment to statistical rationality’ (2022: 93) in which to approach a social or political problem objectively is to be willing to endorse any conclusion, no matter how morally or philosophically questionable – such as in contemporary facial recognition's replication of physiognomy and biological essentialism (Agüera y Arcas et al, 2017; Stark and Hutson, 2022). Datafication's own troubled relationship with the temptation of epistemic certainty facilitates practices of fact signalling and nostalgia, through the common currency of idealised facts as reservoirs of certainty – indeed, facts that ‘don’t care about your feelings’.…”

Section: Mythologising the Factual Pastmentioning

confidence: 99%

Fact signalling and fact nostalgia in the data-driven society

Hong

2023

Big Data & Society

View full text Add to dashboard Cite

Post-truth tells the story of a public descending into unreason, aided and abetted by platforms and other data-driven systems. But this apparent collapse of epistemic consensus is, I argue, also dominated by loud and aggressive commitment to the idea of facts and Reason – a site where an imagined modern past is being pillaged for vestigial legitimacy. This article identifies two common practices of such reappropriation and mythologisation. (1) Fact signalling involves performative invocations of facts and Reason, which are then weaponised to discredit communicative rivals and establish affective solidarity. This is often closely tied to (2) fact nostalgia: the cultivation of an imagined past when ‘facts were facts’ and we, the good liberal subjects, could recognise facts when we saw them. Both tendencies are underwritten by a myth of connection: the still enduring narrative that maximising the circulation of information regardless of provenance or meaning will eventually yield a more rational public – even as data-driven systems tend to undermine the very conditions for such a public. Drawing on examples from YouTube-amplified ‘alternative influencers’ in the American right, and the normative discourses around fact-checking practices, I argue that this continued reliance on the vestigial authority of the modern past is a pernicious obstacle in normative debates around data-driven publics, keeping us stuck on the same dead-end scripts of heroically suspicious individuals and ignorant, irrational masses.

show abstract

Anti-fascist AI

Cited by 17 publications

References 0 publications

Opening up ChatGPT: Tracking openness, transparency, and accountability in instruction-tuned text generators

Opening up ChatGPT: Tracking openness, transparency, and accountability in instruction-tuned text generators

Imaginaries and problematisations: A heuristic lens in the age of artificial intelligence in education

Fact signalling and fact nostalgia in the data-driven society

Contact Info

Product

Resources

About