Datum: 29 Maj, 2023 Nivå: Magisteruppsats i företagsekonomi, 15 hp Institution: Akademin för Ekonomi, Samhälle och Teknik, Mälardalens Universitet Författare: Alan Touma (98/01/12), Shalaw Salih (00/04/03) Titel: Från traditionella arbetsuppgifter till en allt mer rådgivande roll - En kvalitativ undersökning om redovisningskonsulter kompetenskrav i den rådgivande rollen och deras upplevelse av eventuella förväntningsgap Handledare: Dariusz Osowski Nyckelord: Redovisning, rådgivning, automatisering, artificiell intelligens, kompetenser, förväntningsgap Forskningsfråga/or:  • På vilket sätt upplever redovisningskonsulter att deras yrkesroll har påverkats av automatisering och artificiell intelligens? • Vilka kompetenser efterfrågas i den rådgivande rollen och hur skiljer sig dessa från de kompetenser som tidigare har identifierats? • Upplever redovisningskonsulter eventuella förväntansgap som rådgivande redovisningskonsulter och vilka eventuella strategier kan det innebära för att möta kundens behov? Syfte: Syftet med denna studie är att undersöka hur automatisering och artificiell intelligens påverkar redovisningskonsulternas yrkesroll, vilka kompetenser som eftersöks i den rådgivande rollen och hur dessa skiljer sig från tidigare identifierade kompetenser. Dessutom kommer studien att undersöka om redovisningskonsulterna upplever eventuellt förväntningsgap i den rådgivande rollen. Metod: Studien har tillämpat en kvalitativ forskningsmetod med ett abduktiv tillvägagångssätt. Primärdata samlades via semistrukturerade intervjuer med sex utvalda redovisningskonsulter. Den insamlade empirin har sedan analyserats genom tillämpning av tematisk analys. Slutsats: Studien visar att automatisering och artificiell intelligens har effektiviserat redovisningskonsulters arbetsuppgifter. Detta frigör tid för redovisningskonsulter att fokusera på mer analytiska och rådgivande uppgifter. För att framgångsrikt tillfredsställa kundernas behov i denna rådgivande roll behövs kompetenser som social kompetens, analytisk kompetens, teknisk kompetens, specialiserad kunskap samt grundläggande kunskaper. Vidare indikerar studien att förväntningsgapet mellan kunder och redovisningskonsulter inte är betydande. Detta beror på de effektiva strategier som konsulterna använder för att minska detta gap såsom tydlig kommunikation och transparens. / Date: May 29th, 2023 Level: Master thesis in Business Administration, 15 ECTS Institution: School of Business, Society and Engineering, Mälardalen University Authors: Alan Touma (98/01/12), Shalaw Salih (00/04/03) Title: From traditional tasks to an increasingly advisory role - A qualitative study on the competence requirements of accounting consultants in the advisory role and their experience of potential expectation gap Supervisor: Dariusz Osowski Keywords: Accounting, advisory, automation, artificial intelligence,  competencies, expectation gap Research questions:  • In what ways do accounting consultants feel that their professional role has been impacted by automation and artificial intelligence? • What competencies are in demand in the advisory role and how do these differ from the competencies that have been previously identified? • Do accounting consultants experience any expectation gaps as advisory accounting consultants and what possible strategies might this entail to meet customer needs? Purpose: The purpose of this study is to examine how automation and artificial intelligence affect the role of accounting consultants, what competencies are sought in the advisory role, and how these differ from previously identified competencies. In addition, the study will investigate whether accounting consultants perceive a possible expectation gap in the advisory role. Method: The study has applied a qualitative research method with an abductive approach. Primary data were collected through semi structured interviews with six selected accounting consultants. The collected empirical data was then analyzed using thematic analysis. Conclusion: The study shows that automation and artificial intelligence have made accounting consultants' tasks more efficient. This frees up time for accounting consultants to focus on more analytical and advisory tasks. In order to successfully satisfy the customers' needs in this advisory role, competencies such as social competence, analytical competence, technical competence, specialist knowledge and basic knowledge are needed. Furthermore, the study indicates that the expectation gap between customers and accounting consultants is not significant. This is due to the effective strategies that the consultants use to reduce this gap, such as clear communication and transparency.

The AI-based chatbot ChatGPT has taken the world by storm, especially the education sector, as it has the ability to generate full reports, write scripts, assess reports and much more. The disruptive technology ChatGPT is changing the way students learn and the way lecturers teach by offering an exciting new approach to teaching and learning. As ChatGPT has had rapid development, a need for knowledge about the technology has been formed in higher education. The idea of ​​this study is to fill that need and give higher education better knowledge of what ChatGPT has to offer. This study investigates the potential impact of ChatGPT, an artificial intelligence-based algorithm, on the practices and routines in higher education. The study has a qualitative approach and was conducted through three interviews with teachers from different institutions at Umeå University. We also conducted a document study to gather input from teachers across the country to get a wider perspective. The results suggest that the emergence of ChatGPT has the potential to change the way teachers approach their teaching and interact with students. ChatGPT can provide students with quick access to information and feedback, and assist teachers in providing personalized learning experiences and support to students. Additionally, we found that it is important to be critical of the responses generated by ChatGPT since it generates answers based on a series of guesses, which can also be biased. Due to ChatGPTs capabilities of generating realistic texts, institutions may also have to modify examinations and assessment methods to ensure academic integrity. Overall, the findings of this study suggest that the emergence of ChatGPT has the potential to bring significant changes to the practices and routines in higher education, and that teachers should be prepared to adapt to these changes in order to effectively leverage the benefits of this technology.

Background: The technological development has contributed to an increased implementation and use of AI in organizations. Since AI contributes to both opportunities and risks, leaders need knowledge of how intelligent systems can be used. AI and leadership is an unexplored phenomenon that needs to be investigated more closely, in order for leaders to be able to keep up with our rapidly digitized world. Purpose: The purpose of the thesis is to increase understanding of how implementation and use of AI in organizations affects leadership. Furthermore, the thesis aims to find out how leaders can benefit from AI in their role as well as navigate its challenges. Method: The data has been collected through qualitative interviews of semi-structured form. Using an inductive approach, the cross-sectional study treats AI and leadership as a phenomenon.  Conclusion: The role of a leader is changing with the implementation and use of AI in organizations. When AI takes over tasks or facilitates the work of a leader, the central part of leadership becomes the human, relational and social interactions. The positive effects identified are efficiency, improvement and that AI facilitates rather than replaces human resources. Risks with implementing AI in organizations have been identified as ethical dilemmas, ignorance and resistance. / Bakgrund: Den teknologiska utvecklingen har bidragit till en ökad implementering och användning av AI i organisationer. Eftersom AI bidrar med såväl möjligheter som risker behöver ledare kunskap om hur intelligenta system kan användas. AI och ledarskap är ett outforskat fenomen som behöver undersökas närmare för att ledare ska kunna följa med i vår snabbt digitaliserade värld.  Syfte: Uppsatsens syfte är att öka förståelsen för hur implementering och användning av AI i organisationer påverkar ledarskap. Vidare syftar uppsatsen till att ta reda på hur ledare kan dra nytta av AI i sin roll samt navigera dess utmaningar.  Metod: Datan har samlats in genom kvalitativa intervjuer av semistrukturerad form. Med ett induktivt tillvägagångssätt behandlar tvärsnittsstudien AI och ledarskap som ett fenomen.   Slutsats: Rollen som ledare förändras i och med implementering och användning av AI i organisationer. När AI tar över arbetsuppgifter eller underlättar ledarens arbete, blir den centrala delen av ledarskapet det mänskliga, relationella och sociala interaktioner. De positiva effekter som identifierats är effektivisering, förbättring och att AI underlättar snarare än ersätter mänsklig arbetskraft. Risker med implementering av AI i organisationer har identifierats vara etiska dilemman, okunskap och motstånd.

Slutsatsen av denna studie är att en mängd varierande faktorer påverkar en konsuments attityd, både vad gäller artificiell intelligens (AI) men också attityder gentemot företagen som brukar den sagda teknologin. I enlighet med The Technology Acceptance Model (Davies, 1989) är konsumentens uppfattning om huruvida AI personligen kommer gynna dem av yttersta vikt. Såvida ingen vinning uppfattas kommer attityden gentemot AI påverkas negativt. Ytterligare har även faktorer som riskbenägenhet, annonsrelevans och subjektiva normer betydelse i vad som är en konsuments attityd gentemot AI och företag. / The conclusion of this study is that a number of varying elements are affecting the consumer attitude, against artificial intelligence (AI) –but also towards the specific companies who utilize AI. Corresponding with The Technology Acceptance Model (Davies, 1989), the perception of whether AI will benefit the consumer or not is essential. Should no gain be perceived, the attitude towards AI will be affected negatively. Furthermore, aspects such as risk beliefs, ad relevance and subjective norms are elements to consider regarding consumer attitude towards AI and subsequently retail companies.

Bakgrund: Sjukdomen KOL är ett allvarligt sjukdomstillstånd som innebär andnöd, nedsatt livskvalitet och begränsad fysisk kapacitet. Det är en vanlig folksjukdom i Sverige och globalt kräver sjukdomen ungefär 3 miljoner liv varje år. Det är vanligt att få tillfälliga försämringar i sin KOL i form av exacerbationer och fler än 75 % av patienterna drabbas av detta. Ofta leder exacerbationer till högre risk för en sämre sjukdomsprognos samt att dö i förtid. Syfte: Syftet var att beskriva patienters upplevelse av KOL-exacerbation. Metod: Litteraturöversikt med kvalitativ metod och induktiv ansats. Efter kvalitetsgranskning inkluderades tolv artiklar. Dataanalysen genomfördes med Fribergs analysmodell för kvalitativ metoddesign. Resultat: Resultatet omfattade tre huvudteman och åtta subteman. Huvudtemat Informatik med tillhörande subteman; Att bli lyssnad på, Att bli informerad och Bekräftelse. Huvudtemat Autonomi med subteman; Förlust av fysiska aspekter, Förlust av makt och kontroll och tredje huvudtemat Inre Emotionella Processer med tillhörande subteman; Ångest och oro, Rädsla samt Skuld och skam.  Slutsatser: KOL-exacerbation är ett komplext sjukdomstillstånd som orsakar både fysiska och psykiska påfrestningar. Det finns därav utmaningar att vårda patienten ur ett helhetsperspektiv. Personcentrerad vård kan bidra till att patientens behov tillfredsställs även vid KOL-exacerbation. / Background: The disease COPD is a serious illness with shortness of breath, reduced quality of life and limited physical capacity. It is a common disease in Sweden as well as globally. Globally the disease claims approximately 3 million lives each year. It is common to have temporary worsening of their COPD called exacerbations and more than 75% of patients suffer from this. Exacerbations often lead to a higher risk of a worse prognosis of the disease and a premature death. Aim: The experience of exacerbation in chronic obstructive pulmonary disease - from a patient perspective. Methods: A literature review with qualitative method and inductive approach. Twelve articles were included after a quality review. The data analysis was conducted using Friberg's analysis model for qualitative method design. Results: The result comprised three main themes and eight subthemes. The main theme Informatics with subthemes: Being listened to, Being informed and Acknowledgment. The main theme Autonomy with subthemes: Loss of physical aspects, Loss of power and control and the third main theme Inner emotional processes with subthemes: Anxiety and worry, Fear and Guilt and shame.  Conclusions: COPD exacerbation has a negative affect on both the physical and mental health and implies a complex medical condition. There are challenges in caring for the patient from a holistic perspective. Person-centered care can help to meet the patient's needs even during an COPD exacerbation.

Deep neural networks (DNNs) based on satellite and radar data have shown promising results for precipitation nowcasting, beating physical models and optical flow for time horizons up to 8 hours. “MetNet”, developed by Google AI, is a 225 million parameter DNN combining three different types of architectures that was trained on satellite and radar data over the United States. They claim to be the first machine learning model to outperform physical models at such a scale. In this work, we implemented a similar but simplified model trained on radar-only Swedish data, with the aim to perform precipitation nowcasting for up to 2 hours into the future. Furthermore, we compare the model to another, simpler model that omits the spatial aggregator of the DNN architecture which is a state-of-the-art vision transformer. Our results show that, although the adopted training dataset was too small to prevent overfitting, the model is still able to outperform the persistence benchmark for lead times longer than 30 minutes with a threshold of 0.2mm/h precipitation. Our simplified model, perhaps unsurprisingly, is outperformed by MetNet because of having too few training data samples or variances in the models’ implementation. We show, nonetheless, that the adopted spatial aggregator fulfills a vital role as expected, aggregating global information into spatial and temporal contexts.  Due to the limitations imposed by the reduced size of the model, we cannot, unfortunately, draw definitive conclusions on whether a radar-only model could yield similar forecast skills as MetNet. To improve on these results, more training data is certainly needed. This would require that more robust computation resources are available, but pre-training the model on a larger dataset — or even implementing a model that takes in different geographical locations for training — can naturally lead to significant improvements in the predictions. / Djupa neurala nätverk (DNN) baserade på satellit och radar data har gett bra resultat för korta nederbördsprognoser och kan slå fysikaliska modeller och optical flow f ̈or prognoser upp till 8 timmar i framtiden. “MetNet” ̈ar ett 225 million DNN publicerat av Google som kombinerar tre olika typer av djupa arkitekturer, det är tränat på satellit och radar data över USA och är enligt dom den första maskininlärningsmodellen som presterar bättre än fysikaliska modeller. I denna uppsats har vi konstruerat en modell som liknar deras på ett nedskalat problem. Vi har färre parametrar, lägre upplöst data, endast 2 timmar prognostisering och använder bara radar data över Sverige för att träna modellen. Vi använder F1-score för att evaluera modellens prestanda och jämför prognosen mot persistens som referens. Vidare undersöker vi en mindre komplicerad modell där den tredje arkitekturen inte används för att se vilken roll vision transformern har. Våra resultat visar att datasetet vi tränat på är för litet och modellen överanpassas men modellen lyckas ändå slå persistens referensen för prognoser 30–120 minuter när en 0.2mm/h regntröskel tillämpas. Resultaten är sämre än MetNet av Google och vi kan inte dra några slutsatser huruvida en modell med endast radar-data skulle kunna ge liknande resultat eller inte, eftersom modellen inte tränats till dess fulla potential. Vi visar att den tredje arkitekturen, vision transformern, är en viktig del av nätverket och aggregerar global information till lokala kontexter över tid och rum. För att förbättra våra resultat skulle vi pröva att låta modellen träna på det amerikanska datasetet använt av Google och implementera en modell vars input varierar geografisk position.

Idag är artificiell intelligens inom sjukvården ett omtalat ämne. Tidigare litteratur visar att det finns en tydlig potential för AI inom sjukvården där det kan göra stor nytta. Syftet med denna studien har varit att belysa var den svenska sjukvården befinner sig idag och identifiera de begränsningar som finns och försvårar sjukvårdens arbete med AI. Studien har fokuserat på att tala med experter inom området för att få deras perspektiv på situationen.  Empiriska materialet tyder på att den svenska sjukvården inte är tillräckligt förberedd och digitalt mogen för någon större utbreddhet av AI idag. Det påpekas att det krävs en digital transformation där digitala rutiner och processer blir en större del av den svenska sjukvårdens vardag först. Det visar sig att regelverken kring AI och datahantering inte är kompatibla med de ambitioner som finns vad gäller AI inom sjukvården. Denna uppsats visar att det finns ett tydligt glapp mellan förväntningarna kring AI och det faktiska arbetet i sjukvården där optimismen inte är lika tydlig. Det krävs en tydligare nationell samordning, större digital kompetens och framför allt en större ödmjukhet inför att det tar tid att bli digitalt mogen för en sådan stor förändring som AI. / Today, AI in healthcare is a popular topic. Previous literature shows that there is a clear potential for AI in healthcare where it can be of great benefit. The purpose of this study has been to shed light on where Swedish healthcare is today and identify the limitations that exist and complicate healthcare's work with AI. The study has focused on talking to experts in the field to get their perspective on the situation. Empirical evidence suggests that Swedish healthcare is not sufficiently prepared and digitally mature for any greater prevalence of AI today. It is pointed out that a digital transformation is required where digital routines and processes become a larger part of Swedish healthcare's everyday life first. It turns out that the regulations regarding AI and data management are not compatible with the ambitions that exist with regard to AI in healthcare. This essay shows that there is a clear gap between the expectations around AI and the actual work in healthcare where the optimism is not as clear. Clearer national coordination, greater digital competence and, above all, greater humility are required before it takes time to become digitally mature for such a major change as AI.

This study examines how new technologies, particularly generative artificial intelligence (AI), have become an innovative tool for communicators and how it affects their productivity and creativity. Through a combination of literature review, theoretical frameworks, and empirical research, we analyze how communicators integrate generative AI into their work process and how this affects their workflow and conditions. The study is based on the following questions: How are software-based text chat robots used by communicators in their professional role; To what extent do communicators perceive that there is an impact on creativity and productivity when co-writing between human and machine; What opportunities and challenges do communicators imagine that software-based text chat robots can contribute to? The findings indicate that generative AI has the potential to transform the communications industry by increasing efficiency and freeing up time for more strategic thinking and creativity. Communicators report increased productivity and that generative AI has the ability to help communicators manage large text bases in an agile way. At the same time, the study also identifies challenges and potential risks with the use of generative AI. Among these challenges are issues related to ethics, quality assurance and the need to maintain human control and creative input in the creation process. Communicators are aware of these challenges and emphasize the importance of balancing automation with human skills and insight. Finally, the study highlights the opportunities and challenges of the use of generative AI for communicators and identifies areas for future research and development. By understanding the potential benefits and limitations of this technology, communicators can develop strategies to maximize its positive effects and manage its challenges effectively.

This study aims to investigate how communication specialists perceive the integration of generative artificial intelligence (AI) within their profession and how these changes affect their professional role. The purpose of the study was achieved by conducting qualitative interviews with eight communication professionals regarding their view on the integration of generative AI within their profession. Particularly, we found it interesting to examine how the communication specialists use generative AI within their professional role, what risks and/or opportunities they identify concerning the use of generative AI tools within the communication profession and what factors they deem as crucial for successfully and effectively integrating AI tools within their professional roles. The theoretical framework of the study consists of diffusion theory, actor-network theory, risk society and reflexive modernity.  The results of the study indicate that communication specialists see several opportunities in integrating generative AI tools into their work. Key opportunities identified included using generative AI for inspiration and automating repetitive tasks, thereby allowing communication professionals to allocate more time to the strategic aspects of their work. There was no expressed concern that generative AI could currently replace the communication specialist’s professional role. Thus, the risks identified did not seem to revolve around the job security of the communication specialists but rather around the uncertainty about how generative AI functions, its capabilities, and how it handles the input data. Therefore, the establishment of clearer guidelines and policies on AI usage was something that the communication specialists in this study expressed a great need for. Additional factors identified by communication specialists as crucial for successful AI integration included individual characteristics such as curiosity, a willingness to learn, and openness to change. Thus, the results of the study indicate that achieving successful and effective integration of generative AI is perceived to be a shared responsibility at both organizational and individual levels.

In this thesis a new concept called domain knowledge management for informationproviding dialogue systems is introduced. Domain knowledge management includes issues related to representation and use of domain knowledge as well as access of background information sources, issues that previously have been incorporated in dialogue management. The work on domain knowledge management reported in this thesis can be divided in two parts. On a general theoretical level, knowledge sources and models used for dialogue management, including domain knowledge management, are studied and related to the capabilities they support. On a more practical level, domain knowledge management is examined in the contexts of a dialogue system framework and a specific instance of this framework, the ÖTRAF system. In this system domain knowledge management is implemented in a separate module, a Domain Knowledge Manager. The use of a specialised Domain Knowledge Manager has a number of advantages. The first is that dialogue management becomes more focused as it only has to consider dialogue phenomena, while domain-specific reasoning is handled by the Domain Knowledge Manager. Secondly, porting of a system to new domains is facilitated since domain-related issues are separated out in specialised domain knowledge sources. The third advantage with a separate module for domain knowledge management is that domain knowledge sources can be easily modified, exchanged, and reused. / <p>Report code: LiU-Tek-Lic-2001:27.</p>