Digitization and Artificial Intelligence are currently one of the most discussed topics in the real estate industry and are claimed to represent a fundamental revolution rather than a short-term trend or hype. Historically, the real estate industry has been slow to adopt new technology, including traditional analytical AI. However, since OpenAI launched ChatGPT in November 2022, there has been a resurgence of interest in the potential of large language models and more specifically generative AI. The new attention has sparked great enthusiasm for the possibilities offered by generative AI, but also concern about falling behind in its use. The real estate sector is one of the sectors expected to see the absolute largest productivity gains per employee if generative AI is effectively implemented, and with the new applications that traditional AI has not addressed, generative AI potentially offers a unique opportunity for the real estate industry to become a leader in technological development. Despite the promising potential of generative AI, research into its opportunities and challenges in the real estate context is still in its infancy, with a limited number of studies exploring the area. The purpose of this study was thus to contribute to the research on generative AI within the context of property management with the goal of providing insights to Swedish property managers about the socio-technical factors that influence the acceptance of generative AI as well as its potential applications within the industry. Through a survey with 172 participants active in property management, driving forces and obstacles linked to the acceptance of generative AI as well as practical areas of use were investigated. The study's findings highlight the factors that lead a user to either accept or reject generative AI. Also highlighted are the tasks in which property managers see generative AI as a complement, can be performed without or with minimal human supervision, and cannot contribute with support within. Results indicate a wide range of tasks that can potentially be supported by generative AI and are thus valuable to property managers with both limited experience with the technology and those with more experience. / Digitalisering och Artificiell intelligens är för närvarande ett av de mest diskuterade ämnena inom fastighetsbranschen och påstås representera en grundläggande revolution snarare än en kortsiktig trend eller hype. Historiskt sett har fastighetsbranschen varit långsam med att anamma ny teknologi, inklusive traditionell analytisk AI. Sedan OpenAI lanserade ChatGPT i november 2022 har det dock skett ett uppsving i intresset för potentialen hos stora språkmodeller och mer specifikt generative AI. Den nya uppmärksamheten har väckt stor entusiasm över de möjligheter som generativ AI erbjuder, men också en oro över att hamna efter i användningen av den. Fastighetssektorn är en av de sektorerna som förväntas se de absolut största produktivitetsökningarna per anställd om generativ AI implementeras effektivt, och med de nya tillämpningarna som traditionell AI inte har hanterat erbjuder generativ AI potentiellt en unik möjlighet för fastighetsbranschen att bli ledande inom teknisk utveckling. Trots den lovande potentialen hos generativ AI är forskningen om dess möjligheter och utmaningar inom fastighetskontexten fortfarande i begynnelsen, med ett begränsat antal studier som utforskar området. Syftet med denna studie var därmed att bidra till forskningen om generativ AI inom kontexten av fastighetsförvaltning med målet att ge insikter till svenska fastighetsförvaltare om de sociotekniska faktorer som påverkar acceptansen av generativ AI samt dess potentiella tillämningar inom branschen. Genom en enkätundersökning med 172 deltagare verksamma inom fastighetsförvaltning undersöktes drivkrafter och hinder kopplat till acceptans av generativ AI samt praktiska användningsområden. Studiens resultat belyser de faktorer som leder till att en användare antingen accepterar eller avvisar generativ AI. Även belyses de arbetsuppgifter som fastighetsförvaltare ser generativ AI fungera som ett komplement i, kan utföras utan eller med minimal mänsklig övervakning samt inte kan bidra med stöd inom. Resultatet indikerar ett bredd spektrum av arbetsuppgifter som potentiellt kan stödjas av generativ AI och därmed är värdefulla för fastighetsförvaltare med både begränsad erfarenhet av tekniken samt de med mer erfarenhet.

This study explores the integration of ChatGPT into a UX design process, focusing specifically on the creation of user onboarding tutorials. Collaborating with Nasdaq and one of their trade surveillance products, provides a real-world test case. Using A/B testing, the study compares two clickable prototypes: one developed through a conventional process and the other with the assistance of ChatGPT. The objective is twofold: to investigate how ChatGPT can be used to create onboarding tutorials and to evaluate its influence on resource utilization, user experience, and designer workflows. The study concludes that both traditional and ChatGPT-assisted design produces satisfactory results, with no statistically significant difference in knowledge retention or resource utilization. ChatGPT emerges as a valuable UX tool, offering efficiency and versatility in design tasks. However, its effectiveness relies on designers' discernment to shape output into the final user experience. While ChatGPT enhances productivity and creativity, it does not replace human judgment. Yet, leveraging AI tools like ChatGPT can potentially change UX design practices for the better, affirming that AI will not replace UX designers but UX designers leveraging AI will.

This study explores the motivations, challenges, and design opportunities associated with using ChatGPT. The research employs an user-centred design approach to understand user interactions with ChatGPT and propose design concepts. Key motivations for using ChatGPT include its practical utility, ability to provide personalized answers, assistive capabilities, and role as an idea-sparring partner. However, users face challenges such as navigating large amounts of text, understanding how to prompt effectively, and dealing with ChatGPT’s lack of nuanced understanding. Consequently, this project proposes a redesign incorporating interactive features and Graphical User Interface changes to tackle these challenges. The findings suggest that the proposed concepts could significantly improve navigation and glanceability and facilitate the overviewing of past interactions. This research contributes to the field of interaction design by providing insights into the use of conversational generative AI and suggesting improvements for future applications.

This study aims to explore the relationship between AI-enabled enterprises and regulatory compliance, particularly in the integration of Generative AI in auditing and accounting firms. It identifies challenges and provides actionable recommendations for firms adopting AI while ensuring legal compliance. Additionally a qualitative research methodology is used with a thematic analysis approach to explore the integration of Generative AI in auditing and accounting firms in Sweden. Data are collected through semi-structured interviews with four professionals from large auditing firms. This method allows for an in-depth examination of the implications of AI adoption drawing from firsthand insights to identify emerging patterns and themes. Discovering that Swedish auditing and accounting firms adopt Generative AI technologies driven by efficiency gains and enhanced service delivery. This strategic move requires balancing efficiency with regulatory compliance, emphasizing alignment with laws and internal rules. AI development frameworks guide firms in ethical governance, supported by continuous training for staff. Ultimately, this strategic approach ensures both efficiency and ethical integrity in AI adoption.

This thesis has investigated what impact generative AI may have on higher education. Using a combination of a systematic literature study and interviews with representatives from four (4) large universities in Sweden. The findings indicate that generative AI is already a disruptive technology in teaching and learning in higher education, and that students now more easily can cheat or “mislead the examiner” using generative AI, for example by presenting ChatGPT generated text as text written by the students themselves. Even though there are some negatives with generative AI, this thesis shows that the Universities are better off embracing this technology instead of trying to work against it. So, what are the positives with generative AI in education? The fact that students can now converse with someone no matter their background, the fact that students can learn by using ChatGPT (if they are taught how to use it properly), the fact that learning how to use ChatGPT might increase the student’s efficiency and therefore increase their attractiveness on the work market when graduating. All of these benefits come with a big WARNING though. That warning is that higher education must teach the students that these tools are not miracle workers. That the tools can be wrong, and it is important that students learn how to question and criticise what is generated. Higher education has a responsibility to introduce the tools tempered by the understanding that they are not a replacement for knowledge, but only a powerful aid to enhance the knowledge that the students already possess. Finally, the study has been conducted during a particularly expansive period for generative AI and the reader should realise that the findings within this thesis represent early results in a young area of research.

Kolavskiljning, avlägsnande av koldioxid (CO2) från atmosfären, har fått uppmärksamhet som en metod för att mildra effekterna av den globala uppvärmningen. Växter och fototrofa mikroorganismer har den inneboende förmåganatt fånga upp kol genom fixering av CO2 för att producera biomassa. Däremot inhemska kolfixeringsvägar begränsas av nyckelenzymer med låg katalytisk aktivitet vilket resulterar i låg energieffektivitet. Rubisco är en sådan nyckelenzym, ökänt för sin dåliga prestanda. Tidigare forskning har misslyckats när det gäller att förbättra kolet fixering i Rubisco med konventionella metoder. Generativ modellering har dykt upp som en innovativ förhållningssätt till enzymteknik, dra fördel av olika arkitekturer för neurala nätverk för att föreslå en ny varianter med önskade egenskaper. Här tränas en variationsautokodare (VAE) på Rubisco-sekvensen utrymme användes för utmaningen med Rubiscos ingenjörskonst. Två modeller utbildades och med hjälp av dimensionsreduktionsegenskapen hos VAE, utforskades fitnesslandskapet i Rubisco. Sekvenser var märkt med katalytiskt relevanta data och en regressionsmodell byggdes med syftet att förutsäga dessa sekvenser med ökad katalytisk aktivitet. Nya Rubisco-sekvenser genererades efter systematiska utfrågning av det lågdimensionella rummet. Användningen av generativ modellering här ger ett nytt perspektiv på Rubisco engineering. / Carbon capture, the removal of carbon dioxide (CO2) from the atmosphere, has gained attention as a method to mitigate the effects of global warming. Plants and phototrophic microorganisms have the inherent ability to capture carbon through the fixation of CO2 to produce biomass. However, native carbon fixing pathways are limited by key enzymes with low catalytic activity resulting in low energy efficiency. Rubisco is one such key enzyme, notorious for its poor performance. Past research has been unsuccessful at enhancing carbon fixation in Rubisco through conventional methods. Generative modelling has emerged as an innovative approach to enzyme engineering, taking advantage of different neural network architectures to propose novel variants with desired characteristics. Here, a variational autoencoder (VAE) trained on the Rubisco sequence space was applied to the challenge of Rubisco engineering. Two models were trained and, using the dimensionality reduction property of VAEs, the fitness landscape of Rubisco was explored. Sequences were labelled with catalytically relevant data and a regression model was built with the aim of predicting those sequences with enhanced catalytic activity. Novel Rubisco sequences were generated following systematic interrogation of the low-dimensional space. The use of generative modelling here provides a fresh perspective on Rubisco engineering.

Humans perceive and interact with their environment through a multitude of sensory channels. Among these, hearing plays a pivotal role, enabling humans to effectively navigate their surroundings. Sound localization, a complex process, relies on the ability of the human brain to distinguish subtle differences between propagated sound waves interacted with the subject's anthropometric features. However, when utilizing headphones in virtual environments, this natural interaction between sound waves and the human subject is altered. To replicate this phenomenon, the acquisition of head-related filters (HR filters) is necessary to transform non-spatial audio into its spatial representation. Unfortunately, the recording process of HR filters is arduous and resource-intensive, resulting in spatial gaps within datasets, particularly in regions above and below the subject, which are more challenging to capture. To address these incomplete HR filters, extrapolation methods must be employed. While distance extrapolation has been previously explored, research on extrapolation techniques for HR filters remains scarce. Hence, this study introduces a novel approach utilizing a pre-trained deep learning model known as Stable Diffusion to efficiently train the model. The results of this innovative technique showcase a remarkable level of precision and fidelity in the extrapolation of head-related filters (HR filters) for both high and low elevations for virtual auditory environments. Through the utilization of the proposed approach, HR filters are successfully extended beyond their original recording boundaries, allowing for an enhanced spatial representation of sound sources situated at varying heights. The extrapolation process not only achieves high levels of accuracy but also ensures the preservation of intricate spatial details, enabling a more immersive and realistic auditory experience for users. These findings signify a significant advancement in the field of virtual acoustics and hold substantial implications for applications such as virtual reality, gaming, and audio engineering. / Människor uppfattar och interagerar med sin omgivning genom en mängd sensoriska kanaler. Bland dessa har hörseln en avgörande roll och möjliggör för människor att effektivt navigera i sin omgivning. Ljudlokalisering, en komplex process, är beroende av människans förmåga att urskilja subtila skillnader mellan interagerande ljudvågor med människans antropometriska särdrag. När dock hörlurar används i virtuella miljöer förändras denna naturliga interaktion mellan ljudvågor och människan. För att replikera detta fenomen behövs insamling av huvudrelaterade filter (HR-filter) för att omvandla icke-spatialt ljud till dess spatiala representation. Tyvärr är inspelningsprocessen för HR-filter besvärlig och resurskrävande, vilket resulterar i spatiala luckor inom datamängder, särskilt i områden över och under subjektet som är svårare att fånga. För att åtgärda dessa ofullständiga HR-filter måste extrapolationsmetoder användas. Medan avståndsextrapolation tidigare har undersökts är forskningen kring extrapolationstekniker för HR-filter knapphändig. Därför presenterar denna studie ett nytt tillvägagångssätt som utnyttjar en förtränad djupinlärningsmodell kallad Stable Diffusion för att effektivt träna modellen. Resultaten från denna innovativa teknik visar en anmärkningsvärd precision och noggrannhet vid extrapoleringen av huvudrelaterade filter (HR-filter) för både höga och låga höjdpositioner för virtuella ljudmiljöer. Genom användning av det föreslagna tillvägagångssättet kan HR-filter framgångsrikt förlängas utanför sina ursprungliga inspelningsgränser, vilket möjliggör en förbättrad spatial representation av ljudkällor som är placerade på olika höjder. Extrapolationsprocessen uppnår inte bara hög noggrannhet utan säkerställer också bevarandet av intrikata spatiala detaljer, vilket möjliggör en mer immersiv och realistisk ljudupplevelse för användarna. Dessa resultat innebär en betydande framsteg inom området virtuell akustik och har väsentliga implikationer för tillämpningar såsom virtuell verklighet, spel och ljudteknik.

Denna studie undersöker användningen av kodassistenter som bygger på generativ AI (GAI) för systemutveckling och testning inom myndigheter och större företag, med fokus på säkerhetsrisker och funktionella krav. Detta görs genom en kartläggning där intervjuer och interna såväl som publika dokument används som datainsamlingsmetod. Med den snabba utvecklingen av stora språkmodeller (LLM:s) och deras tillämpningar inom systemutveckling är det viktigt för organisationer att göra val av AI-verktyg som uppfyller både säkerhets- och funktionskrav. Studien kartlägger och jämför olika verktyg baserat på dessa kriterier, med syfte att hjälpa organisationer att göra välavvägda val i ett ständigt växande utbud av GAIverktyg för programmering. Säkerhetsriskerna som identifierats inkluderar risken för informationsläckor och generering av osäker kod. Studien föreslår mitigerande åtgärder som användning av verktyg som inte sparar eller tränar på användardata för att minska dessa risker. Funktionella krav som identifierats inkluderar förmågan att generera och förklara kod, samt integration med utvecklingsmiljöer som Visual Studio och VisualStudio Code. Verktygen filtrerades därefter baserat på de identifierade mitigeringarna, vilket resulterade i en rekommendation av sex verktyg som uppfyller säkerhetskraven. Fyra av dessa sex verktyg uppfyller i sin tur de ställda kriterierna från samarbetspartnern (Trafikverket) gällande funktionskrav. Denna studie bidrar till den akademiska forskningen genom att erbjuda en ökad förståelse för säkerhets- och funktionsaspekterna av GAI-verktyg (ChatGPT, Claude, Codeium, Codiumate, Cody, GitHub Copilot, Gemini, Q, Tabnine) inom systemutveckling och testning. Detta genom en lista över potentiella säkerhetsrisker och en interaktiv jämförelsetabell där identifierade GAI-verktygkan filtreras efter olika krav. / This study examines the use of generative AI (GAI) coding assistants for system development and testing within government agencies and big companies, focusing on security risks and functional requirements. This is done through a survey where interviews and both internal and public documents are used as datacollection methods. With the rapid development of large language models (LLMs) and their growing use in system development, it is crucial for organizations to choose AI tools that     meet both security and functionality requirements. This survey compares various tools based on these criteria, aiming to assist organizations in making wellbalanced choices in an ever-growing range of GAI tools for coding. The security risks identified include the risk of information leaks and the generation of insecure code. The study suggests mitigations such as the use of tools that do not save or train on user data to reduce these risks. Functional requirements identified include the ability to generate and explain code, as well as integration with popular development environments like Visual Studio and Visual Studio Code. The tools were then filtered based on the identified mitigations, resulting in a recommendation of six tools that meet the security requirements. Four of these six tools, in turn, meet the set criteria from the partner (SwedishTransport Administration) regarding functional requirements. This study contributes to academic research by offering an increased understanding of the security and functionality aspects of GAI tools (ChatGPT, Claude, Codeium, Codiumate, Cody, GitHub Copilot, Gemini, Q, Tabnine) within system development and testing. This through a list of potential security risks and an interactive comparison table where identified GAI tools can be filtered according to different requirements.

The integration of generative AI in the design industry has gained attention for its potential to enhance productivity and spark creativity. However, there is a gap in how to effectively incorporate the technology into the early stages of the design process. To address this, interviews were conducted with digital product designers at a consultancy agency to understand their current approaches to early-stage design and attitudes toward using generative AI tools into it. Findings from the interviews were used to inform the development of use case scenarios of generative AI tools in early design, which was later evaluated with the participants through observational interviews using ChatGPT and Midjourney. These findings indicate that digital product designers view generative AI as a valuable tool that can streamline the design process, aid in research and exploration, support communication and documentation, and inspire creativity. However, concerns were raised regarding trustworthiness, reliability, biased training data, and limited creativity and human nuance in these models. Designers believe in using AI-generated output as a guide, support, or inspiration, rather than a substitute for their own creativity and critical thinking, which enables an approach that allows for a mindful but effective interaction with generative AI tools.

he specter of deepfakes and artificial intelligence enabled media productioncontinues to exacerbate the fear brought on by a degraded ability to discern the real from the fake, syn- thetic, or fabricated in a networked society. While these fears are well-founded especially as they pertain to issues of involuntary pornography their introduction into an already oversaturated media landscape, if anything, extended trends in mediated indeterminacy already being fostered by the universalization of social media platforms. Sites such as Facebook, Instagram, Snapchat and TikTok, made more explicit the contingency and per- formative nature of identity. As younger generations came of age through social media they learned how to navigate and present themselves through it in novel ways unique to each platform. Oftentimes, these strategies were harmful to people’s perception of themselves and their mental health. Other times, however, it gave them the ability to experiment with new forms of identity more in line with how they actually felt. Further, more experimentation through ubiquitous mediation extended what kinds of identities are possible in general as well. In turn, the discovery and extension of identity has led to the evolution of desire. Identities and desires hitherto not possible in a physical space precipitated the creation of new objects of desire that can be pursued and materially experienced regardless of their virtual nature. Deepfakes, and now generative AI, anticipate a further, exponentially more complicated relationship with identity and desire formation through the adoption of increasingly unreal presentations of each. / Media Studies & Production