Pub Date : 2024-03-14DOI: 10.1109/TTS.2024.3377541
Emily J. Zuetell;Paulina Jaramillo
Data-driven policy development and investment are necessary for aligning policies across administrative levels, targeting interventions, and meeting the 2030 Sustainable Development Goals. However, local-level economic well-being data at timely intervals, critical to informing policy development and ensuring equity of outcomes, are unavailable in many parts of the world. Yet, filling these data gaps with black-box predictive models like neural networks introduces risk and inequity to the decision- making process. In this work, we construct an alternative interpretable model to these black-box models to predict household wealth, a key socioeconomic well-being indicator, at 5-km scale from widely available satellite data. Our interpretable model promotes transparency, the identification of potential drivers of bias and harmful outcomes, and inclusive design for human-ML decision-making. We model household wealth as a low- order function of productive land use that can be interpreted and integrated with domain knowledge by decision-makers. We aggregate remotely sensed land cover change data from 2006–2019 to construct an interpretable linear regression model for household wealth and wealth change in Uganda at a 5-km scale with $r^{2},,{=}$ 72%. Our results demonstrate that there is not a clear performance-interpretability tradeoff in modeling household wealth from satellite imagery at high spatial and temporal resolution. Finally, we recommend a tiered framework to model socioeconomic outcomes from remote sensing data.
{"title":"A Framework for the Interpretable Modeling of Household Wealth in Rural Communities From Satellite Data","authors":"Emily J. Zuetell;Paulina Jaramillo","doi":"10.1109/TTS.2024.3377541","DOIUrl":"https://doi.org/10.1109/TTS.2024.3377541","url":null,"abstract":"Data-driven policy development and investment are necessary for aligning policies across administrative levels, targeting interventions, and meeting the 2030 Sustainable Development Goals. However, local-level economic well-being data at timely intervals, critical to informing policy development and ensuring equity of outcomes, are unavailable in many parts of the world. Yet, filling these data gaps with black-box predictive models like neural networks introduces risk and inequity to the decision- making process. In this work, we construct an alternative interpretable model to these black-box models to predict household wealth, a key socioeconomic well-being indicator, at 5-km scale from widely available satellite data. Our interpretable model promotes transparency, the identification of potential drivers of bias and harmful outcomes, and inclusive design for human-ML decision-making. We model household wealth as a low- order function of productive land use that can be interpreted and integrated with domain knowledge by decision-makers. We aggregate remotely sensed land cover change data from 2006–2019 to construct an interpretable linear regression model for household wealth and wealth change in Uganda at a 5-km scale with \u0000<inline-formula> <tex-math>$r^{2},,{=}$ </tex-math></inline-formula>\u0000 72%. Our results demonstrate that there is not a clear performance-interpretability tradeoff in modeling household wealth from satellite imagery at high spatial and temporal resolution. Finally, we recommend a tiered framework to model socioeconomic outcomes from remote sensing data.","PeriodicalId":73324,"journal":{"name":"IEEE transactions on technology and society","volume":"5 1","pages":"36-44"},"PeriodicalIF":0.0,"publicationDate":"2024-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141164734","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-03-11DOI: 10.1109/TTS.2024.3375519
Kausik Lakkaraju;Biplav Srivastava;Marco Valtorta
Sentiment Analysis Systems (SASs) are data-driven Artificial Intelligence (AI) systems that assign one or more numbers to convey the polarity and emotional intensity of a given piece of text. However, like other automatic machine learning systems, SASs can exhibit model uncertainty, resulting in drastic swings in output with even small changes in input. This issue becomes more problematic when inputs involve protected attributes like gender or race, as it can be perceived as bias or unfairness. To address this, we propose a novel method to assess and rate SASs. We perturb inputs in a controlled causal setting to test if the output sentiment is sensitive to protected attributes while keeping other components of the textual input, such as chosen emotion words, fixed. Based on the results, we assign labels (ratings) at both fine-grained and overall levels to indicate the robustness of the SAS to input changes. The ratings can help decision-makers improve online content by reducing hate speech, often fueled by biases related to protected attributes such as gender and race. These ratings provide a principled basis for comparing SASs and making informed choices based on their behavior. The ratings also benefit all users, especially developers who reuse off-the-shelf SASs to build larger AI systems but do not have access to their code or training data to compare.
情感分析系统（SAS）是一种数据驱动的人工智能（AI）系统，它可以分配一个或多个数字来表达给定文本的极性和情感强度。然而，与其他自动机器学习系统一样，SAS 也会表现出模型的不确定性，导致即使输入的微小变化也会导致输出的剧烈波动。当输入涉及性别或种族等受保护的属性时，这个问题就变得更加棘手，因为这可能被视为偏见或不公平。为了解决这个问题，我们提出了一种评估和评价 SAS 的新方法。我们在受控的因果关系设置中对输入进行扰动，以测试输出情感是否对受保护属性敏感，同时保持文本输入的其他成分（如所选情感词）固定不变。根据结果，我们在细粒度和整体层面上分配标签（评级），以显示 SAS 对输入变化的鲁棒性。这些评级可以帮助决策者通过减少仇恨言论来改进在线内容，而仇恨言论往往是由性别和种族等受保护属性的偏见所助长的。这些评级为比较 SAS 和根据其行为做出明智选择提供了原则性依据。这些评级也有利于所有用户，特别是那些重复使用现成的 SAS 构建大型人工智能系统，但却无法获得其代码或训练数据进行比较的开发人员。
{"title":"Rating Sentiment Analysis Systems for Bias Through a Causal Lens","authors":"Kausik Lakkaraju;Biplav Srivastava;Marco Valtorta","doi":"10.1109/TTS.2024.3375519","DOIUrl":"https://doi.org/10.1109/TTS.2024.3375519","url":null,"abstract":"Sentiment Analysis Systems (SASs) are data-driven Artificial Intelligence (AI) systems that assign one or more numbers to convey the polarity and emotional intensity of a given piece of text. However, like other automatic machine learning systems, SASs can exhibit model uncertainty, resulting in drastic swings in output with even small changes in input. This issue becomes more problematic when inputs involve protected attributes like gender or race, as it can be perceived as bias or unfairness. To address this, we propose a novel method to assess and rate SASs. We perturb inputs in a controlled causal setting to test if the output sentiment is sensitive to protected attributes while keeping other components of the textual input, such as chosen emotion words, fixed. Based on the results, we assign labels (ratings) at both fine-grained and overall levels to indicate the robustness of the SAS to input changes. The ratings can help decision-makers improve online content by reducing hate speech, often fueled by biases related to protected attributes such as gender and race. These ratings provide a principled basis for comparing SASs and making informed choices based on their behavior. The ratings also benefit all users, especially developers who reuse off-the-shelf SASs to build larger AI systems but do not have access to their code or training data to compare.","PeriodicalId":73324,"journal":{"name":"IEEE transactions on technology and society","volume":"5 1","pages":"82-92"},"PeriodicalIF":0.0,"publicationDate":"2024-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141164678","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Digital personal data are increasingly understood as a key asset in our digital economies. But how should we value such data? Numerous policymakers, regulators, and stakeholders are trying to work out how to manage the collection, use, and valuation of data in order to balance the advantages and disadvantages of its collection and use. The negative implications of data practices may include privacy loss, data breaches, or declining market competition, while social and economic benefits include improved service delivery, more efficient welfare systems, or better products. Increasingly, data are conceptualized as an asset. To understand the value of data as an asset means understanding how data are configured as an asset; data value does not reflect ownership and property rights per se, but rather diverse modes of access and use restrictions (usually delineated by opaque contractual agreements). Data are increasingly controlled by a few, large digital technology firms, especially so-called ‘Big Tech’ firms. In this paper, we use Google as a case study of how Big Tech firms configure and value digital data as an asset. We analyse how Google understands, frames, values, and monetizes the data they collect from users. We qualitatively analyse an extensive dataset of financial documentary materials produced by and about Google to identify the different modes of access and use restrictions that Google deploys to turn digital data into a valuable asset. We conclude that, despite being highly ambiguous, Google’s approach to data value focuses on monetizing users.
{"title":"Understanding Data Valuation: Valuing Google’s Data Assets","authors":"Kean Birch;Sarah Marquis;Guilherme Cavalcante Silva","doi":"10.1109/TTS.2024.3398400","DOIUrl":"https://doi.org/10.1109/TTS.2024.3398400","url":null,"abstract":"Digital personal data are increasingly understood as a key asset in our digital economies. But how should we value such data? Numerous policymakers, regulators, and stakeholders are trying to work out how to manage the collection, use, and valuation of data in order to balance the advantages and disadvantages of its collection and use. The negative implications of data practices may include privacy loss, data breaches, or declining market competition, while social and economic benefits include improved service delivery, more efficient welfare systems, or better products. Increasingly, data are conceptualized as an asset. To understand the value of data as an asset means understanding how data are configured as an asset; data value does not reflect ownership and property rights per se, but rather diverse modes of access and use restrictions (usually delineated by opaque contractual agreements). Data are increasingly controlled by a few, large digital technology firms, especially so-called ‘Big Tech’ firms. In this paper, we use Google as a case study of how Big Tech firms configure and value digital data as an asset. We analyse how Google understands, frames, values, and monetizes the data they collect from users. We qualitatively analyse an extensive dataset of financial documentary materials produced by and about Google to identify the different modes of access and use restrictions that Google deploys to turn digital data into a valuable asset. We conclude that, despite being highly ambiguous, Google’s approach to data value focuses on monetizing users.","PeriodicalId":73324,"journal":{"name":"IEEE transactions on technology and society","volume":"5 2","pages":"183-190"},"PeriodicalIF":0.0,"publicationDate":"2024-03-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10525235","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141964835","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-03-07DOI: 10.1109/TTS.2024.3374639
Razvan Amironesei;Mark Díaz
The development of many machine learning (ML) and artificial intelligence (AI) systems depends on human-labeled data. Human-provided labels act as tags or enriching information that enable algorithms to more easily learn patterns in data in order to train or evaluate a wide range of AI systems. These annotations ultimately shape the behavior of AI systems. Given the scale of ML datasets, which can contain thousands to billions of data points, cost and efficiency play a major role in how data annotations are collected. Yet, important challenges arise between the goals of meeting scale-related needs while also collecting data in a way that reflects real-world nuance and variation. Annotators are typically treated as interchangeable workers who provide a ‘view from nowhere’. We question assumptions of universal ground truth by focusing on the social and ethical aspects that shape annotation task design.
{"title":"Social and Ethical Norms in Annotation Task Design","authors":"Razvan Amironesei;Mark Díaz","doi":"10.1109/TTS.2024.3374639","DOIUrl":"https://doi.org/10.1109/TTS.2024.3374639","url":null,"abstract":"The development of many machine learning (ML) and artificial intelligence (AI) systems depends on human-labeled data. Human-provided labels act as tags or enriching information that enable algorithms to more easily learn patterns in data in order to train or evaluate a wide range of AI systems. These annotations ultimately shape the behavior of AI systems. Given the scale of ML datasets, which can contain thousands to billions of data points, cost and efficiency play a major role in how data annotations are collected. Yet, important challenges arise between the goals of meeting scale-related needs while also collecting data in a way that reflects real-world nuance and variation. Annotators are typically treated as interchangeable workers who provide a ‘view from nowhere’. We question assumptions of universal ground truth by focusing on the social and ethical aspects that shape annotation task design.","PeriodicalId":73324,"journal":{"name":"IEEE transactions on technology and society","volume":"5 1","pages":"45-47"},"PeriodicalIF":0.0,"publicationDate":"2024-03-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141164724","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-03-07DOI: 10.1109/TTS.2024.3374647
Bijun Wang;Onur Asan;Ting Liao;Mo Mansouri
Artificial intelligence (AI) can transform various aspects of healthcare, including diagnosis, treatment, monitoring, and preventative care. Patients’ attitudes and views are considered critical factors for the development and success of AI-based technology in healthcare delivery. This study seeks to explore the chronic patients’ perceptions, including their knowledge, concerns regarding misuse and abuse, their attitude toward AI involvement, and their views on the future role of AI in healthcare delivery. Using the convenience sampling technique, 219 chronic-condition participants completed an online survey. This study leveraged Hayes Process Macro to develop a moderated mediation model to analyze the collected data. Our results showed that patients’ knowledge of AI did not directly influence their perceptions of the future of AI in healthcare. Nonetheless, the evidence from the mediational analysis revealed an indirect effect, where concerns about AI misuse and abuse and extensive AI involvement played a role in that. Additionally, the level of trustworthiness moderated the relationship between acceptance of extensive AI involvement and patients’ perception of AI’s future role. These findings highlight the importance of considering patients’ views and attitudes towards AI and addressing any concerns or fears they may have in order to build trust and confidence in clinical AI systems, which can ultimately lead to better health outcomes.
人工智能（AI）可以改变医疗保健的各个方面，包括诊断、治疗、监测和预防性护理。患者的态度和观点被认为是人工智能技术在医疗保健领域的发展和成功的关键因素。本研究旨在探讨慢性病患者的看法，包括他们的知识、对误用和滥用的担忧、他们对人工智能参与的态度以及他们对人工智能在医疗保健服务中未来角色的看法。通过方便抽样技术，219 名慢性病患者完成了在线调查。本研究利用 Hayes Process Macro 建立了一个调节中介模型来分析收集到的数据。我们的研究结果表明，患者对人工智能的了解程度并不直接影响他们对人工智能在医疗保健领域未来发展的看法。不过，中介分析的证据显示了一种间接影响，即对人工智能误用和滥用的担忧以及人工智能的广泛参与在其中发挥了作用。此外，可信赖程度调节了接受人工智能广泛参与与患者对人工智能未来作用的看法之间的关系。这些研究结果突显了考虑患者对人工智能的看法和态度以及解决他们可能存在的任何担忧或恐惧的重要性，从而建立对临床人工智能系统的信任和信心，最终实现更好的医疗效果。
{"title":"The Future Role of Clinical Artificial Intelligence: View of Chronic Patients","authors":"Bijun Wang;Onur Asan;Ting Liao;Mo Mansouri","doi":"10.1109/TTS.2024.3374647","DOIUrl":"https://doi.org/10.1109/TTS.2024.3374647","url":null,"abstract":"Artificial intelligence (AI) can transform various aspects of healthcare, including diagnosis, treatment, monitoring, and preventative care. Patients’ attitudes and views are considered critical factors for the development and success of AI-based technology in healthcare delivery. This study seeks to explore the chronic patients’ perceptions, including their knowledge, concerns regarding misuse and abuse, their attitude toward AI involvement, and their views on the future role of AI in healthcare delivery. Using the convenience sampling technique, 219 chronic-condition participants completed an online survey. This study leveraged Hayes Process Macro to develop a moderated mediation model to analyze the collected data. Our results showed that patients’ knowledge of AI did not directly influence their perceptions of the future of AI in healthcare. Nonetheless, the evidence from the mediational analysis revealed an indirect effect, where concerns about AI misuse and abuse and extensive AI involvement played a role in that. Additionally, the level of trustworthiness moderated the relationship between acceptance of extensive AI involvement and patients’ perception of AI’s future role. These findings highlight the importance of considering patients’ views and attitudes towards AI and addressing any concerns or fears they may have in order to build trust and confidence in clinical AI systems, which can ultimately lead to better health outcomes.","PeriodicalId":73324,"journal":{"name":"IEEE transactions on technology and society","volume":"5 1","pages":"71-81"},"PeriodicalIF":0.0,"publicationDate":"2024-03-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141164722","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-03-01DOI: 10.1109/TTS.2024.3370095
Jordan Richard Schoenherr;Robert Thomson
While previous studies of trust in artificial intelligence have focused on perceived user trust, the paper examines how an external agent (e.g., an auditor) assigns responsibility, perceives trustworthiness, and explains the successes and failures of AI. In two experiments, participants (university students) reviewed scenarios about automation failures and assigned perceived responsibility, trustworthiness, and preferred explanation type. Participants’ cumulative responsibility ratings for three agents (operators, developers, and AI) exceeded 100%, implying that participants were not attributing trust in a wholly rational manner, and that trust in the AI might serve as a proxy for trust in the human software developer. Dissociation between responsibility and trustworthiness suggested that participants used different cues, with the kind of technology and perceived autonomy affecting judgments. Finally, we additionally found that the kind of explanation used to understand a situation differed based on whether the AI succeeded or failed.
{"title":"When AI Fails, Who Do We Blame? Attributing Responsibility in Human–AI Interactions","authors":"Jordan Richard Schoenherr;Robert Thomson","doi":"10.1109/TTS.2024.3370095","DOIUrl":"https://doi.org/10.1109/TTS.2024.3370095","url":null,"abstract":"While previous studies of trust in artificial intelligence have focused on perceived user trust, the paper examines how an external agent (e.g., an auditor) assigns responsibility, perceives trustworthiness, and explains the successes and failures of AI. In two experiments, participants (university students) reviewed scenarios about automation failures and assigned perceived responsibility, trustworthiness, and preferred explanation type. Participants’ cumulative responsibility ratings for three agents (operators, developers, and AI) exceeded 100%, implying that participants were not attributing trust in a wholly rational manner, and that trust in the AI might serve as a proxy for trust in the human software developer. Dissociation between responsibility and trustworthiness suggested that participants used different cues, with the kind of technology and perceived autonomy affecting judgments. Finally, we additionally found that the kind of explanation used to understand a situation differed based on whether the AI succeeded or failed.","PeriodicalId":73324,"journal":{"name":"IEEE transactions on technology and society","volume":"5 1","pages":"61-70"},"PeriodicalIF":0.0,"publicationDate":"2024-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141164703","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper serves as an Introduction to the Public Interest Technology (PIT) for Innovation in Global Development Special Issue, based on a workshop of the same title held in September 2023. The paper’s contribution is in proposing recommendations and practical guidance to aid in launching PIT projects. We begin by situating the Special Issue in evolving definitions of PIT in Section II , followed by an overview of the PIT ecosystem in Section III to offer a succinct account of the current state of PIT scholarship. The corresponding links to the innovation in global development context are subsequently described in Section IV , in keeping with the theme of the workshop. These links relate to an overview of adjacent fields and concepts; an illustrative example in the information technology for development (ICT4D) field; the identification of gaps in current PIT scholarship; and the preliminary questions that require attention. Next, Section V presents workshop outcomes, in the form of a general overview of the event; the identification of prevalent themes emerging from and / or are reinforced in the workshop; and a summary of Special Issue papers. The workshop is used as an interdisciplinary catalyst for the explication of more recent PIT developments. These developments are encapsulated in ten recommendations for launching PIT projects in Section VI , intended to direct PIT project managers or lead investigators prior to project launch or during the initial stages of a project.
{"title":"Public Interest Technology for Innovation in Global Development: Recommendations for Launching PIT Projects","authors":"Roba Abbas;Katina Michael;Dinara Davlembayeva;Savvas Papagiannidis;Jeremy Pitt","doi":"10.1109/TTS.2024.3375431","DOIUrl":"https://doi.org/10.1109/TTS.2024.3375431","url":null,"abstract":"This paper serves as an Introduction to the Public Interest Technology (PIT) for Innovation in Global Development Special Issue, based on a workshop of the same title held in September 2023. The paper’s contribution is in proposing recommendations and practical guidance to aid in launching PIT projects. We begin by situating the Special Issue in evolving definitions of PIT in \u0000<xref>Section II</xref>\u0000, followed by an overview of the PIT ecosystem in \u0000<xref>Section III</xref>\u0000 to offer a succinct account of the current state of PIT scholarship. The corresponding links to the innovation in global development context are subsequently described in \u0000<xref>Section IV</xref>\u0000, in keeping with the theme of the workshop. These links relate to an overview of adjacent fields and concepts; an illustrative example in the information technology for development (ICT4D) field; the identification of gaps in current PIT scholarship; and the preliminary questions that require attention. Next, \u0000<xref>Section V</xref>\u0000 presents workshop outcomes, in the form of a general overview of the event; the identification of prevalent themes emerging from and / or are reinforced in the workshop; and a summary of Special Issue papers. The workshop is used as an interdisciplinary catalyst for the explication of more recent PIT developments. These developments are encapsulated in ten recommendations for launching PIT projects in \u0000<xref>Section VI</xref>\u0000, intended to direct PIT project managers or lead investigators prior to project launch or during the initial stages of a project.","PeriodicalId":73324,"journal":{"name":"IEEE transactions on technology and society","volume":"5 1","pages":"14-23"},"PeriodicalIF":0.0,"publicationDate":"2024-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10539349","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141164723","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-03-01DOI: 10.1109/TTS.2024.3378587
Katina Michael;Jordan Richard Schoenherr;Kathleen M. Vogel
The dark side of AI has been a persistent focus in discussions of popular science and academia (Appendix A), with some claiming that AI is “evil” [1] . Many commentators make compelling arguments for their concerns. Techno-elites have also contributed to the polarization of these discussions, with ultimatums that in this new era of industrialized AI, citizens will need to “[join] with the AI or risk being left behind” [2] . With such polarizing language, debates about AI adoption run the risk of being oversimplified. Discussion of technological trust frequently takes an all-or-nothing approach. All technologies – cognitive, social, material, or digital – introduce tradeoffs when they are adopted, and contain both ‘light and dark’ features [3] . But descriptions of these features can take on deceptively (or unintentionally) anthropomorphic tones, especially when stakeholders refer to the features as ‘agents’ [4] , [5] . When used as an analogical heuristic, this can inform the design of AI, provide knowledge for AI operations, and potentially even predict its outcomes [6] . However, if AI agency is accepted at face value, we run the risk of having unrealistic expectations for the capabilities of these systems.
{"title":"Failures in the Loop: Human Leadership in AI-Based Decision-Making","authors":"Katina Michael;Jordan Richard Schoenherr;Kathleen M. Vogel","doi":"10.1109/TTS.2024.3378587","DOIUrl":"https://doi.org/10.1109/TTS.2024.3378587","url":null,"abstract":"The dark side of AI has been a persistent focus in discussions of popular science and academia (Appendix A), with some claiming that AI is “evil” \u0000<xref>[1]</xref>\u0000. Many commentators make compelling arguments for their concerns. Techno-elites have also contributed to the polarization of these discussions, with ultimatums that in this new era of industrialized AI, citizens will need to “[join] with the AI or risk being left behind” \u0000<xref>[2]</xref>\u0000. With such polarizing language, debates about AI adoption run the risk of being oversimplified. Discussion of technological trust frequently takes an \u0000<italic>all-or-nothing</i>\u0000 approach. All technologies – cognitive, social, material, or digital – introduce tradeoffs when they are adopted, and contain both ‘light and dark’ features \u0000<xref>[3]</xref>\u0000. But descriptions of these features can take on deceptively (or unintentionally) anthropomorphic tones, especially when stakeholders refer to the features as ‘agents’ \u0000<xref>[4]</xref>\u0000, \u0000<xref>[5]</xref>\u0000. When used as an analogical heuristic, this can inform the design of AI, provide knowledge for AI operations, and potentially even predict its outcomes \u0000<xref>[6]</xref>\u0000. However, if AI agency is accepted at face value, we run the risk of having unrealistic expectations for the capabilities of these systems.","PeriodicalId":73324,"journal":{"name":"IEEE transactions on technology and society","volume":"5 1","pages":"2-13"},"PeriodicalIF":0.0,"publicationDate":"2024-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10539317","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141164676","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-03-01DOI: 10.1109/TTS.2024.3386247
Ehsan Nabavi;Rob Nicholls;George Roussos
Whether we, as end-users of technology, are aware of it or not, our societies are becoming increasingly entangled in a complex network of interactions with Artificial Intelligence (AI) systems. This goes beyond what is often called ‘Human-AI collaboration’, involving the broader socio-political systems supporting these technologies [1] . Faced with these complex interactions, society grapples with the timeless question: where does responsibility lie for the consequences or results produced by AI systems or applications, whether they are successful or not? Is it with the human operator, AI developer, user, or the AI agent itself? In the case of failure, AI cannot be held accountable, as such software systems are not yet recognized as separate legal entities [2] .
{"title":"Locating Responsibility in the Future of Human–AI Interactions","authors":"Ehsan Nabavi;Rob Nicholls;George Roussos","doi":"10.1109/TTS.2024.3386247","DOIUrl":"https://doi.org/10.1109/TTS.2024.3386247","url":null,"abstract":"Whether we, as end-users of technology, are aware of it or not, our societies are becoming increasingly entangled in a complex network of interactions with Artificial Intelligence (AI) systems. This goes beyond what is often called ‘Human-AI collaboration’, involving the broader socio-political systems supporting these technologies \u0000<xref>[1]</xref>\u0000. Faced with these complex interactions, society grapples with the timeless question: where does responsibility lie for the consequences or results produced by AI systems or applications, whether they are successful or not? Is it with the human operator, AI developer, user, or the AI agent itself? In the case of failure, AI cannot be held accountable, as such software systems are not yet recognized as separate legal entities \u0000<xref>[2]</xref>\u0000.","PeriodicalId":73324,"journal":{"name":"IEEE transactions on technology and society","volume":"5 1","pages":"58-60"},"PeriodicalIF":0.0,"publicationDate":"2024-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10539358","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141164721","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-03-01DOI: 10.1109/TTS.2024.3374980
{"title":"IEEE Transactions on Technology and Society Publication Information","authors":"","doi":"10.1109/TTS.2024.3374980","DOIUrl":"https://doi.org/10.1109/TTS.2024.3374980","url":null,"abstract":"","PeriodicalId":73324,"journal":{"name":"IEEE transactions on technology and society","volume":"5 1","pages":"C2-C2"},"PeriodicalIF":0.0,"publicationDate":"2024-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10539307","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141164679","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","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：