Pub Date : 2023-01-01DOI: 10.1177/20539517231166886
Rainer Mühlhoff
Big data and artificial intelligence pose a new challenge for data protection as these techniques allow predictions to be made about third parties based on the anonymous data of many people. Examples of predicted information include purchasing power, gender, age, health, sexual orientation, ethnicity, etc. The basis for such applications of “predictive analytics” is the comparison between behavioral data (e.g. usage, tracking, or activity data) of the individual in question and the potentially anonymously processed data of many others using machine learning models or simpler statistical methods. The article starts by noting that predictive analytics has a significant potential to be abused, which manifests itself in the form of social inequality, discrimination, and exclusion. These potentials are not regulated by current data protection law in the EU; indeed, the use of anonymized mass data takes place in a largely unregulated space. Under the term “predictive privacy,” a data protection approach is presented that counters the risks of abuse of predictive analytics. A person's predictive privacy is violated when personal information about them is predicted without their knowledge and against their will based on the data of many other people. Predictive privacy is then formulated as a protected good and improvements to data protection with regard to the regulation of predictive analytics are proposed. Finally, the article points out that the goal of data protection in the context of predictive analytics is the regulation of “prediction power,” which is a new manifestation of informational power asymmetry between platform companies and society.
{"title":"Predictive privacy: Collective data protection in the context of artificial intelligence and big data","authors":"Rainer Mühlhoff","doi":"10.1177/20539517231166886","DOIUrl":"https://doi.org/10.1177/20539517231166886","url":null,"abstract":"Big data and artificial intelligence pose a new challenge for data protection as these techniques allow predictions to be made about third parties based on the anonymous data of many people. Examples of predicted information include purchasing power, gender, age, health, sexual orientation, ethnicity, etc. The basis for such applications of “predictive analytics” is the comparison between behavioral data (e.g. usage, tracking, or activity data) of the individual in question and the potentially anonymously processed data of many others using machine learning models or simpler statistical methods. The article starts by noting that predictive analytics has a significant potential to be abused, which manifests itself in the form of social inequality, discrimination, and exclusion. These potentials are not regulated by current data protection law in the EU; indeed, the use of anonymized mass data takes place in a largely unregulated space. Under the term “predictive privacy,” a data protection approach is presented that counters the risks of abuse of predictive analytics. A person's predictive privacy is violated when personal information about them is predicted without their knowledge and against their will based on the data of many other people. Predictive privacy is then formulated as a protected good and improvements to data protection with regard to the regulation of predictive analytics are proposed. Finally, the article points out that the goal of data protection in the context of predictive analytics is the regulation of “prediction power,” which is a new manifestation of informational power asymmetry between platform companies and society.","PeriodicalId":47834,"journal":{"name":"Big Data & Society","volume":null,"pages":null},"PeriodicalIF":8.5,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"49067969","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-01-01DOI: 10.1177/20539517231182393
Riccardo Pronzato
In recent years, several scholars have highlighted the necessity to scrutinize the practices and material settings in which algorithmic models are designed, in order to unpack the working activities and socio-cultural constructs underlying their production and deployment process. Drawing on a multisited ethnography, this paper investigates the practices of tech workers within the corporate environment of an internet television platform, the hierarchical relationships between different professional figures, and how these individuals frame algorithms and contribute to the enactment of these systems with their activities. Findings highlight the hierarchical organization of tech work and the subordination of operative figures to the goals imposed by business clients and to both internal and external forms of control. Specifically, it emerges how the subalternity of tech workers is materially and discursively constructed and forms of causal, dispositional and facilitative power exerted on them. In this environment, frictions, negotiations as well as concealing strategies by tech workers regarding the design and meaning of algorithms emerge, thus showing their cultural, contingent and multiple composition. Within the framework of Giddens’ structure/agency cycle, it is shown how everyday working activities and relationships contribute to the reproduction of hegemonic arrangements in the workplace, and how these hegemonic arrangements are at the core of algorithmic production, thus playing a key role in the framing, construction and enactment of these systems.
{"title":"Algorithms and hegemony in the workplace: Negotiating design and values in an Italian television platform","authors":"Riccardo Pronzato","doi":"10.1177/20539517231182393","DOIUrl":"https://doi.org/10.1177/20539517231182393","url":null,"abstract":"In recent years, several scholars have highlighted the necessity to scrutinize the practices and material settings in which algorithmic models are designed, in order to unpack the working activities and socio-cultural constructs underlying their production and deployment process. Drawing on a multisited ethnography, this paper investigates the practices of tech workers within the corporate environment of an internet television platform, the hierarchical relationships between different professional figures, and how these individuals frame algorithms and contribute to the enactment of these systems with their activities. Findings highlight the hierarchical organization of tech work and the subordination of operative figures to the goals imposed by business clients and to both internal and external forms of control. Specifically, it emerges how the subalternity of tech workers is materially and discursively constructed and forms of causal, dispositional and facilitative power exerted on them. In this environment, frictions, negotiations as well as concealing strategies by tech workers regarding the design and meaning of algorithms emerge, thus showing their cultural, contingent and multiple composition. Within the framework of Giddens’ structure/agency cycle, it is shown how everyday working activities and relationships contribute to the reproduction of hegemonic arrangements in the workplace, and how these hegemonic arrangements are at the core of algorithmic production, thus playing a key role in the framing, construction and enactment of these systems.","PeriodicalId":47834,"journal":{"name":"Big Data & Society","volume":null,"pages":null},"PeriodicalIF":8.5,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43356507","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-01-01DOI: 10.1177/20539517231160528
Shirley Kempeneer, Frederik Heylen
The users, sensors and networks of the Internet of Things generate huge amounts of data. Given the sophisticated (artificially intelligent) algorithms, computing power and software available, we would expect governments to have successfully completed their digital transformation into Jane Fountain's (2001) ‘Virtual State’. In practice, despite heavy investments, governments often fail to enact new digital technologies in an efficient, appropriate or fair way. This article provides an overview of techno-rational and socio-political failures and solutions at the macro-, meso- and micro-level to support digital transformation. The reviewed articles suggest a modest approach to digital transformation, with an emphasis on high-quality in-house IT infrastructure and expertise, but also better collaborative networks and strong leadership ensuring human oversight.
{"title":"Virtual state, where are you? A literature review, framework and agenda for failed digital transformation","authors":"Shirley Kempeneer, Frederik Heylen","doi":"10.1177/20539517231160528","DOIUrl":"https://doi.org/10.1177/20539517231160528","url":null,"abstract":"The users, sensors and networks of the Internet of Things generate huge amounts of data. Given the sophisticated (artificially intelligent) algorithms, computing power and software available, we would expect governments to have successfully completed their digital transformation into Jane Fountain's (2001) ‘Virtual State’. In practice, despite heavy investments, governments often fail to enact new digital technologies in an efficient, appropriate or fair way. This article provides an overview of techno-rational and socio-political failures and solutions at the macro-, meso- and micro-level to support digital transformation. The reviewed articles suggest a modest approach to digital transformation, with an emphasis on high-quality in-house IT infrastructure and expertise, but also better collaborative networks and strong leadership ensuring human oversight.","PeriodicalId":47834,"journal":{"name":"Big Data & Society","volume":null,"pages":null},"PeriodicalIF":8.5,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42230082","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-01-01DOI: 10.1177/20539517231168101
H. Faxon, Kendra Kintzi, Van Tran, Kay Zak Wine, Swan Ye Htut
Despite perennial hope in the democratic possibilities of the internet, the rise of digital authoritarianism threatens online and offline freedom across much of the world. Yet while critical data studies has expanded its geographic focus, limited work to date has examined digital mobilization in the agrarian communities that comprise much of the Global South. This article advances the concept of “organic online politics,” to demonstrate how digital mobilization grows from specific rural conditions, material concerns, and repertoires of resistance, within the constraints of authoritarian violence and internet control. To do so, we examine social media interaction in the wake of the 2021 military coup in Myanmar, an agrarian nation with recent, rapid digital connection that corresponded with a decade-long democratic turn. Analyzing an original archive of over 2000 Facebook posts collected from popular farming pages and groups, we find a massive drop-off in online activity after the military coup and analyze the shifting temporalities of digital mobilization. Crucially, we highlight the embeddedness of online interaction within the material concerns of farming communities, examining how social media become a key forum for negotiating political crisis in Myanmar's countryside. These findings call attention to rural digital subcultures as fertile sites of investigation and point toward the need for future scholarship on data practices that attends to rooted agrarian struggles.
{"title":"Organic online politics: Farmers, Facebook, and Myanmar's military coup","authors":"H. Faxon, Kendra Kintzi, Van Tran, Kay Zak Wine, Swan Ye Htut","doi":"10.1177/20539517231168101","DOIUrl":"https://doi.org/10.1177/20539517231168101","url":null,"abstract":"Despite perennial hope in the democratic possibilities of the internet, the rise of digital authoritarianism threatens online and offline freedom across much of the world. Yet while critical data studies has expanded its geographic focus, limited work to date has examined digital mobilization in the agrarian communities that comprise much of the Global South. This article advances the concept of “organic online politics,” to demonstrate how digital mobilization grows from specific rural conditions, material concerns, and repertoires of resistance, within the constraints of authoritarian violence and internet control. To do so, we examine social media interaction in the wake of the 2021 military coup in Myanmar, an agrarian nation with recent, rapid digital connection that corresponded with a decade-long democratic turn. Analyzing an original archive of over 2000 Facebook posts collected from popular farming pages and groups, we find a massive drop-off in online activity after the military coup and analyze the shifting temporalities of digital mobilization. Crucially, we highlight the embeddedness of online interaction within the material concerns of farming communities, examining how social media become a key forum for negotiating political crisis in Myanmar's countryside. These findings call attention to rural digital subcultures as fertile sites of investigation and point toward the need for future scholarship on data practices that attends to rooted agrarian struggles.","PeriodicalId":47834,"journal":{"name":"Big Data & Society","volume":null,"pages":null},"PeriodicalIF":8.5,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44149784","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-01-01DOI: 10.1177/20539517231153811
Troels Krarup, Maja Horst
Rapid innovation in digital services relying on artificial intelligence (AI) challenges existing regulations across a wide array of policy fields. The European Union (EU) has pursued a position as global leader on ethical AI regulation in explicit contrast to US laissez-faire and Chinese state surveillance approaches. This article asks how the seemingly heterogeneous approaches of market making and ethical AI are woven together at a deeper level in EU regulation. Combining quantitative analysis of all official EU documents on AI with in-depth reading of key reports, communications, and legislative corpora, we demonstrate that single market integration constitutes a fundamental but overlooked engine and structuring principle of new AI regulation. Under the influence of this principle, removing barriers to competition and the free flow of data, on the one hand, and securing ethical and responsible AI, on the other hand, are seen as compatible and even mutually reinforcing.
{"title":"European artificial intelligence policy as digital single market making","authors":"Troels Krarup, Maja Horst","doi":"10.1177/20539517231153811","DOIUrl":"https://doi.org/10.1177/20539517231153811","url":null,"abstract":"Rapid innovation in digital services relying on artificial intelligence (AI) challenges existing regulations across a wide array of policy fields. The European Union (EU) has pursued a position as global leader on ethical AI regulation in explicit contrast to US laissez-faire and Chinese state surveillance approaches. This article asks how the seemingly heterogeneous approaches of market making and ethical AI are woven together at a deeper level in EU regulation. Combining quantitative analysis of all official EU documents on AI with in-depth reading of key reports, communications, and legislative corpora, we demonstrate that single market integration constitutes a fundamental but overlooked engine and structuring principle of new AI regulation. Under the influence of this principle, removing barriers to competition and the free flow of data, on the one hand, and securing ethical and responsible AI, on the other hand, are seen as compatible and even mutually reinforcing.","PeriodicalId":47834,"journal":{"name":"Big Data & Society","volume":null,"pages":null},"PeriodicalIF":8.5,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48671953","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-01-01DOI: 10.1177/20539517231164120
M. Leese, Silvan Pollozek
In this paper, we investigate the temporal implications of data in law enforcement and border control. We start from the assumption that the velocity of knowledge and action is defined by heterogeneous formations and interactions of various actors, sites, and materials. To analyze these formations and interactions, we introduce and unpack the concept of “data temporality.” Data temporality explicates how the speed of knowledge and action in datafied environments unfolds in close correspondence with (1) variegated social rhythms, (2) technological inscriptions, and (3) the balancing of speed with other priorities. Specifically, we use the notion of data temporality as a heuristic tool to explore the entanglements of data and time within two case studies: Frontex’ Joint Operation Reporting Application and the predictive policing software PRECOBS. The analysis identifies two key themes in the empirical constitution of data temporalities. The first one pertains to the creation of events as reference points for temporally situated knowledge and action. And the second one pertains to timing and actionability, that is, the question of when interventions based on data analysis should be triggered.
{"title":"Not so fast! Data temporalities in law enforcement and border control","authors":"M. Leese, Silvan Pollozek","doi":"10.1177/20539517231164120","DOIUrl":"https://doi.org/10.1177/20539517231164120","url":null,"abstract":"In this paper, we investigate the temporal implications of data in law enforcement and border control. We start from the assumption that the velocity of knowledge and action is defined by heterogeneous formations and interactions of various actors, sites, and materials. To analyze these formations and interactions, we introduce and unpack the concept of “data temporality.” Data temporality explicates how the speed of knowledge and action in datafied environments unfolds in close correspondence with (1) variegated social rhythms, (2) technological inscriptions, and (3) the balancing of speed with other priorities. Specifically, we use the notion of data temporality as a heuristic tool to explore the entanglements of data and time within two case studies: Frontex’ Joint Operation Reporting Application and the predictive policing software PRECOBS. The analysis identifies two key themes in the empirical constitution of data temporalities. The first one pertains to the creation of events as reference points for temporally situated knowledge and action. And the second one pertains to timing and actionability, that is, the question of when interventions based on data analysis should be triggered.","PeriodicalId":47834,"journal":{"name":"Big Data & Society","volume":null,"pages":null},"PeriodicalIF":8.5,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48139171","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-01-01DOI: 10.1177/20539517231158636
T. Deruelle, Veronika Kalouguina, Philipp Trein, J. Wagner
A crucial challenge for personalized health is the handling of individuals’ data and specifically the protection of their privacy. Secure storage of personal health data is of paramount importance to convince citizens to collect personal health data. In this survey experiment, we test individuals’ willingness to produce and store personal health data, based on different storage options and whether this data is presented as common good or private good. In this paper, we focus on the nonmedical context with two means to self-produce data: connected devices that record physical activity and genetic tests that appraise risks of diseases. We use data from a survey experiment fielded in Switzerland in March 2020 and perform regression analyses on a representative sample of Swiss citizens in the French- and German-speaking cantons. Our analysis shows that respondents are more likely to use both apps and tests when their data is framed as a private good to be stored by individuals themselves. Our results demonstrate that concerns regarding the privacy of personal heath data storage trumps any other variable when it comes to the willingness to use personalized health technologies. Individuals prefer a data storage format where they retain control over the data. Ultimately, this study presents results susceptible to inform decision-makers in designing privacy in personalized health initiatives.
{"title":"Designing privacy in personalized health: An empirical analysis","authors":"T. Deruelle, Veronika Kalouguina, Philipp Trein, J. Wagner","doi":"10.1177/20539517231158636","DOIUrl":"https://doi.org/10.1177/20539517231158636","url":null,"abstract":"A crucial challenge for personalized health is the handling of individuals’ data and specifically the protection of their privacy. Secure storage of personal health data is of paramount importance to convince citizens to collect personal health data. In this survey experiment, we test individuals’ willingness to produce and store personal health data, based on different storage options and whether this data is presented as common good or private good. In this paper, we focus on the nonmedical context with two means to self-produce data: connected devices that record physical activity and genetic tests that appraise risks of diseases. We use data from a survey experiment fielded in Switzerland in March 2020 and perform regression analyses on a representative sample of Swiss citizens in the French- and German-speaking cantons. Our analysis shows that respondents are more likely to use both apps and tests when their data is framed as a private good to be stored by individuals themselves. Our results demonstrate that concerns regarding the privacy of personal heath data storage trumps any other variable when it comes to the willingness to use personalized health technologies. Individuals prefer a data storage format where they retain control over the data. Ultimately, this study presents results susceptible to inform decision-makers in designing privacy in personalized health initiatives.","PeriodicalId":47834,"journal":{"name":"Big Data & Society","volume":null,"pages":null},"PeriodicalIF":8.5,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45290123","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-01-01DOI: 10.1177/20539517221149097
Rodrigo De La Fabián, Álvaro Jiménez-Molina, Francisco Pizarro Obaid
This article critically examines the emergence and uses of digital phenotyping in contemporary psychiatry. From an analysis of its discourses and practices, we show that digital phenotyping diffusion is directly related to its promise to solve some of the major impasses of the so-called "neuro-turn" in contemporary psychiatry. However, more than a new tool to address old objects of pre-digital psychiatry, we consider digital phenotyping as participating from a new onto-epistemological matrix, the “neuro-digital complex,” which entails the redefinition of psychiatric objects (e.g., brain and mind), diagnostic categories and procedures, subjectivities (e.g., users of mental health apps), and the emergence of a new regime of truth which promises to reveal the neuropsychological core at the individual scale. Despite this techno-utopia, digital phenotyping does not produce neutral mirrors for self-knowledge. We show that it resorts to population statistics, grounded truth data sets built with pre-digital neuropsychological assumptions, and human categorization processes. Nevertheless, we propose not to approach this gap as a misleading ideological fact but to emphasize its productive possibilities. From this perspective, the gap becomes the measure between whom we think we are and who we really are, working as a guide to conduct our lives in neuropsychological terms. Thus, we conclude that, rather than providing personalized diagnoses and treatments, digital phenotyping produces individualized pathways to normalization and neuropsychologization.
{"title":"A critical analysis of digital phenotyping and the neuro-digital complex in psychiatry","authors":"Rodrigo De La Fabián, Álvaro Jiménez-Molina, Francisco Pizarro Obaid","doi":"10.1177/20539517221149097","DOIUrl":"https://doi.org/10.1177/20539517221149097","url":null,"abstract":"This article critically examines the emergence and uses of digital phenotyping in contemporary psychiatry. From an analysis of its discourses and practices, we show that digital phenotyping diffusion is directly related to its promise to solve some of the major impasses of the so-called \"neuro-turn\" in contemporary psychiatry. However, more than a new tool to address old objects of pre-digital psychiatry, we consider digital phenotyping as participating from a new onto-epistemological matrix, the “neuro-digital complex,” which entails the redefinition of psychiatric objects (e.g., brain and mind), diagnostic categories and procedures, subjectivities (e.g., users of mental health apps), and the emergence of a new regime of truth which promises to reveal the neuropsychological core at the individual scale. Despite this techno-utopia, digital phenotyping does not produce neutral mirrors for self-knowledge. We show that it resorts to population statistics, grounded truth data sets built with pre-digital neuropsychological assumptions, and human categorization processes. Nevertheless, we propose not to approach this gap as a misleading ideological fact but to emphasize its productive possibilities. From this perspective, the gap becomes the measure between whom we think we are and who we really are, working as a guide to conduct our lives in neuropsychological terms. Thus, we conclude that, rather than providing personalized diagnoses and treatments, digital phenotyping produces individualized pathways to normalization and neuropsychologization.","PeriodicalId":47834,"journal":{"name":"Big Data & Society","volume":null,"pages":null},"PeriodicalIF":8.5,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45840469","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-01-01DOI: 10.1177/20539517231169731
Xuenan Cao, Roozbeh Yousefzadeh
The right to artificial intelligence (AI) explainability has consolidated as a consensus in the research community and policy-making. However, a key component of explainability has been missing: extrapolation, which can reveal whether a model is making inferences beyond the boundaries of its training. We report that AI models extrapolate outside their range of familiar data, frequently and without notifying the users and stakeholders. Knowing whether a model has extrapolated or not is a fundamental insight that should be included in explaining AI models in favor of transparency, accountability, and fairness. Instead of dwelling on the negatives, we offer ways to clear the roadblocks in promoting AI transparency. Our commentary accompanies practical clauses useful to include in AI regulations such as the AI Bill of Rights, the National AI Initiative Act in the United States, and the AI Act by the European Commission.
{"title":"Extrapolation and AI transparency: Why machine learning models should reveal when they make decisions beyond their training","authors":"Xuenan Cao, Roozbeh Yousefzadeh","doi":"10.1177/20539517231169731","DOIUrl":"https://doi.org/10.1177/20539517231169731","url":null,"abstract":"The right to artificial intelligence (AI) explainability has consolidated as a consensus in the research community and policy-making. However, a key component of explainability has been missing: extrapolation, which can reveal whether a model is making inferences beyond the boundaries of its training. We report that AI models extrapolate outside their range of familiar data, frequently and without notifying the users and stakeholders. Knowing whether a model has extrapolated or not is a fundamental insight that should be included in explaining AI models in favor of transparency, accountability, and fairness. Instead of dwelling on the negatives, we offer ways to clear the roadblocks in promoting AI transparency. Our commentary accompanies practical clauses useful to include in AI regulations such as the AI Bill of Rights, the National AI Initiative Act in the United States, and the AI Act by the European Commission.","PeriodicalId":47834,"journal":{"name":"Big Data & Society","volume":null,"pages":null},"PeriodicalIF":8.5,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46691380","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-01-01DOI: 10.1177/20539517231182402
P. Brodie
In this article, I take the case of data centers as a powerful tool and infrastructure of multinational digital capitalism, analyzing the ways in which understanding these and other data infrastructures through their energy frameworks allows us to theorize the implications of planetary environmental impacts of digital data for contemporary subjects beyond individual data technologies themselves. This is especially true in data centers’ function as energy vacuums and in their carbon and extractive footprints and other environmental externalities. I demonstrate that data centers organize an assemblage of environmental relations whose operations reproduce uneven systems of capitalism enacted through energy and environmental politics. While this article is by no means comprehensive, and by necessity must be selective in its engagement with key texts in a number of overlapping fields, it broadly draws from media studies, geographical, and sociological approaches to data infrastructures to unravel the entanglements of digital systems and the environment. Data centers and their energy connections represent multivalent sites and indications into the global supply chain of data infrastructure, and their extractive dynamic as networked infrastructure fundamentally changes how we need to see their impacts and the impacts of datafication more broadly.
{"title":"Data infrastructure studies on an unequal planet","authors":"P. Brodie","doi":"10.1177/20539517231182402","DOIUrl":"https://doi.org/10.1177/20539517231182402","url":null,"abstract":"In this article, I take the case of data centers as a powerful tool and infrastructure of multinational digital capitalism, analyzing the ways in which understanding these and other data infrastructures through their energy frameworks allows us to theorize the implications of planetary environmental impacts of digital data for contemporary subjects beyond individual data technologies themselves. This is especially true in data centers’ function as energy vacuums and in their carbon and extractive footprints and other environmental externalities. I demonstrate that data centers organize an assemblage of environmental relations whose operations reproduce uneven systems of capitalism enacted through energy and environmental politics. While this article is by no means comprehensive, and by necessity must be selective in its engagement with key texts in a number of overlapping fields, it broadly draws from media studies, geographical, and sociological approaches to data infrastructures to unravel the entanglements of digital systems and the environment. Data centers and their energy connections represent multivalent sites and indications into the global supply chain of data infrastructure, and their extractive dynamic as networked infrastructure fundamentally changes how we need to see their impacts and the impacts of datafication more broadly.","PeriodicalId":47834,"journal":{"name":"Big Data & Society","volume":null,"pages":null},"PeriodicalIF":8.5,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46179469","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：