|
|
GANs gone wild: Public perceptions of Deepfake technologies on YouTube
Poon, Jessica ; Špelda, Petr (advisor) ; Fitzgerald, James (referee)
Deepfake technologies are a form of artificial intelligence (AI) which are based on generative adversarial networks (GANs), a development which has emerged out of deep learning (DL) and machine learning (ML) models. Using a data range which spans the years 2018 - 2021, this research explores public perceptions of deepfake technologies at scale by closely examining commentary found on the social video-sharing platform, YouTube. This open source, ground-level data documents civilian responses to a selection of user-produced, labelled deepfake content. This research fills a gap regarding public perception of this emerging technology at scale. It gauges an underrepresented set of responses in discourse to find that users demonstrate a spectrum of responses which veer between irony and concern, with greater volumes of commentary skewed towards the former. This study of user commentary also finds that YouTube as a wild space ultimately affords reflexive and critical thinking around the subject of deepfake technologies and could prove to be effective as a form of inoculation against disinformation.
|
|
| |
|
|
Shadow warrior: The outsourcing of the United States' offensive cyber capabilities
Jones, Ashley ; Erkomashvili, David (advisor) ; Fitzgerald, James (referee)
The cyberspace has emerged as a conscientious topic for military leaders and senior policymakers as they attempt to ensure critical national infrastructure is secured against cyberattacks. The previous decade has seen the re-emergence of the state actor as a primary threat actor; a large-scale threat actor that is heavily funded, well equipped, and has the ability to avoid legal ramifications for their actions. The cyberspace presented new opportunities of attack and manoeuvre below the threshold of armed response and has transformed the way conflicts are fought and disputes are resolved. This dissertation aims to analyse the effect that the private sector has on the United States' offensive cyber capabilities. Specifically, this research will answer the question of 'How does the private sector support the development of the United States offensive cyber capabilities?' To meet the growing threat of cyber warfare, new organisational structures were rapidly devised and established within the U.S. Department of Defense. Literature on the two key themes, the cyberspace as a strategic capability and the cyberspace as an outsourced commodity, was reviewed to determine the factors contributing to the increase in outsourcing offensive cyber capabilities. Then a qualitative case study on the newest branch...
|
|
|
Experts, politics, and crisis management: A case study of the French expert crisis in the Covid-19 pandemic
Gannac, Chloé ; Střítecký, Vít (advisor) ; Fitzgerald, James (referee)
The primary focus of this research is the epistemic crisis of the scientific expert community. It explores how the context of Covid-19 highlights a state of epistemic crisis among experts. With a case study design based on the example of France during the first wave of the pandemic, it analyses how mechanisms such as politicization of expertise and expertization of politics can influence a shift of scientific expertise from rational science to the realm of normative politics. This research further examines the necessity of public trust for scientific legitimacy and credibility and investigates the place of scientific rigor on the perception of science's epistemic authority. This work subsequently discusses how the different mechanisms described are symptomatic of the scientific experts' epistemic crisis. Lastly, it opens up the discussion around the question of to what extent is this crisis coherent with the narrative of a broader revisionist crisis of the role of public authority figures and the western liberal democratic model. Keywords: epistemic crisis, expertise, Covid-19, expertization, politicization, France, revisionist crisis.
|
|
|
The correlation between the failure of intelligence structures and conflict: The Cabo Delgado case in Northern Mozambique
Ali, Neyma Mildred Mahomed ; Karásek, Tomáš (advisor) ; Fitzgerald, James (referee)
TITLE: The Correlation Between the Failure of Intelligence Structures and Conflict: The Cabo Delgado Case in Northern Mozambique. SUBTITLE: The Study of Organisational Deficiencies of the Intelligence Agency in Mozambique culminating in Intelligence Failure: The Cabo Delgado Case ABSTRACT This paper addresses the intelligence failure of SISE (the State Information and Security Service- Mozambican intelligence agency which is rooted in the organisational defects of the agency. The case study further establishes how this failure contributed to the protracted Islamic insurgency in Cabo Delgado province by failing to prevent the insurgency. In light of this, this dissertation attempts to fill a gap in the literature, analysing and evaluating intelligence failure in the African context. The theoretical assumptions of organisational failure pursued in Amy Zegart's work highlighting structural malfunctions, cultural pathologies and "perverse" incentives ( Zegart, 2007) with regards to organisational failure will be utilised as the conceptual and theoretical framework as it is the most relevant in evaluating the SISE. These organisational disorders utilised as a structured operational framework in this thesis intend to a) determine what kind of deficiencies of the organisational structure and management are...
|
|
|
Hidden conflict in Nigeria: The escalation of the herder-farmer conflict in Nigeria
Iduma, Ugo Igariwey ; Ludvík, Zdeněk (advisor) ; Fitzgerald, James (referee)
The research explores the escalation of the herder-farmer conflict in Nigeria to identify the significant patterns of escalation. Relying on a mixed-method analysis of secondary data and aligning with the analytical anchorage of dynamic systems theory, the research argues that the although Benue and Enugu observe the same herder-farmer the patterns of conflict escalation is neither similar, linear or recurrent. This research submits ethnoreligious antagonism, lawlessness, and exclusionary politics as reasons why the conflict escalated into widespread violence. Adding that each of these elements self- reinforces and influence each other to sustain a coordinated state of violence or maintain peace. It makes a case for pragmatic policies that captures the history and political, economic, and social interaction of states and local government.
|
|
| |
|
|
Machine learning applications in the United States criminal justice system: A critical content analysis of the COMPAS recidivism risk assessment
Bejarano Carbo, Maria Patricia ; Špelda, Petr (advisor) ; Fitzgerald, James (referee)
Artificial intelligence and machine learning (AI/ML) models are increasingly utilised in every aspect of life and society due to their superhuman abilities to digest large amounts of data and find obscure patterns and correlations. One contentious area of this technological application is in the criminal justice system, where AI/ML is used as a recommendation or decision-making support tool. These applications are particularly popular in the United States of America (USA), the nation with the highest rate of incarceration and correctional budget, to aid in managing overcrowded and overspending facilities. Angwin et al.'s (2016) ground-breaking study found the Correctional Offender Management Profiling for Alternative Sanctions (COMPAS) model to be biased against Black defendants and sparked an influential academic debate around algorithmic bias and fairness. This study aims to fill the gap in the scholarship by focusing on the content of COMPAS's recidivism risk assessment questionnaire through a qualitative content analysis within the conceptual framework of Critical Race Theory (CRT). The findings presented in this research are twofold: (1) almost half of the COMPAS questions were opinion-based, thus reducing quantitative neutrality, and (2) there were significant proxy factors for race that...
|
|
|
Making a terrorist: The discursive construction of Islamist and right-wing extremist threats in Swiss media reporting
Margna, Livia ; Prina, Federica (advisor) ; Fitzgerald, James (referee)
Dissertation｜2460442M Abstract The discourses structuring news coverage of terrorist attacks influence our understanding of the nature, drivers and severity of the threat emanating from a specific extremist actor category. Therefore, they are a powerful tool to further socio-political goals. Acknowledging the role of language in shaping reality, this dissertation project uses Critical Discourse Analysis/Critical Discourse Studies to reveal current discursive trends in the understudied coverage of Islamist and right-wing extremist attacks in the Swiss press. With the dominant social factor distinguishing the two extremist categories being ethnicity, it hypothesises that Western media discourses reflect the presuppositions of Orientalism and Critical Race Theory. Both theories expect texts to express, enact and legitimise social hierarchies based on racial affinity to solidify the supremacy of the white elite. The exemplarily analysis of the reporting of two recent extremist incidents by three newspapers representing political perspectives from the right-wing to the left-wing shows that while the Swiss press is indeed influenced by and reproduces racial inequalities, publications do so to a varying degree.
|
|
|
Rule of the algorithm: Exploring an ill-posed problem for democracy
Hawking, Christie ; Vostal, Filip (advisor) ; Fitzgerald, James (referee)
We are entering a new epoch, the epoch of rule by algorithm. This new way of governing poses significant challenges for the wide-ranging scope of democratic functions and practices. This is through the deployment of complex automated decision-making algorithms (ADA) in public institutions that are changing the way that citizens and governments interact with each other, acting as a systematic medium for making public good entitlement and intervention decisions. Although the technical capacities of ADA systems embed nominal challenges that are directly problematic to democratic practices of accountability and transparency, the current research emphasises that the key problems stem from institutionalised mechanisms and decision- making processes. These highlighted challenges underlie the current debates that situate technological innovation and governance to regime specific qualities of either an autocratic or democratic nature. Moreover, it realises that a complex ideological competition exists between democracy and efficiency, which ties into a discussion on the incompatibility of democracy and capitalism. The research narrows in on the current scope and nature of either phenomena by utilising the United States and the United Kingdom as case studies for contextualisation and situational analysis. It...
|
guest ::
