Studiens syfte är att bedöma till vilken utsträckning AI kan ersätta en människa i rollen som mjukvaruutvecklare utifrån ett kvalitativt perspektiv på kod. Detta görs genom att besvara forskningsfrågorna som lyder: “Hur använder mjukvaruutvecklare sig av generativ AI vid utvecklingsutmaningar?” och “Vad är mjukvaruutvecklares uppfattning om kvaliteten på autogenererad kod skapad av en generativ AI såsom Chat GPT?”. För att besvara frågorna har en kvalitativ metod applicerats. En litteraturundersökning startade studien och tillsammans med en ny modell som baseras på McCall quality model och Boehm quality model. Från detta har en intervjuguide skapats som används i semistrukturerade intervjuer genomförda med erfarna mjukvaruutvecklare. Resultatet visar att kod skapad av generativ AI är ett bra hjälpmedel och verktyg som kan effektivisera en mjukvaruutvecklare och att det används på det sättet idag. Däremot så visar resultaten också att koden som genereras av en generativ AI inte är tillräckligt bra och kan inte användas utan att förändringar eller åtgärder görs då det saknas kvalitet. Slutsatserna som dras är att mjukvaruutvecklare använder sig av generativ AI som ett hjälpmedel men att AI:n inte är kapabel att hantera en uppgift på egen hand, därav är det inget hot mot någon anställning för mjukvaruutvecklare. Framtida forskning bör göras på autogenererad kod. Fler verktyg bör undersökas för att utvidga kunskapen om dess kapacitet samt bör det undersökas vilken inverkan generativ AI kan ha på andra branscher. / The aim of this study, conducted and written in Swedish, is to assess the potential of replacing a human software developer with generative AI. The study evaluates the quality of code generated by a generative AI model, this is done by answering the following research questions: “How do software developers use generative AI for development challenges'' and “How do software developers perceive the quality of code autogenerated by a generative AI such as Chat GPT”. To answer the questions we employ a qualitative research method. The study began with a literature review and based our evaluation of software quality on a hybrid model that modifies and combines McCall quality model and Boehm software quality model. The literature review and the hybrid model was used as a base to shape an interview guide. The interview guide was used in semistructured interviews conducted with experienced software developers. The results suggest that autogenerated code from generative AI is a viable aid for software developers as it makes them more effective in a number of tasks. However, the results also show that the autogenerated AI code has insufficient quality as a complete solution, and therefore often requires further fine-tuning and improvements from software developers. From the results, we conclude that software developers do use generative AI as a tool while writing code. Generative AI enhances software developers effectiveness but the current state of generative AI cannot fully replace a human software developer hence it is not a threat to any employment. Future research should be conducted on auto generated code. Some more tools should be studied to broaden the knowledge on its capabilities as well as looking at the implications that generative AI have on other industries.

Background: In today's society, the rapid growth of online shopping among consumers has resulted in the widespread adoption of artificial intelligence powered chatbots by businesses. Despite this, previous research has not focused on where customers place their attitudes towards AI-powered chatbots in e-commerce. The ABC model of attitudes, which includes affective, behavioral, and cognitive components was used to examine how attitudes are formed towards AI-powered chatbots. Purpose: The purpose of this study is to explore consumers' attitudes towards AI-powered chatbots utilized in e-commerce. Methodology: Because the study has an exploratory and inductive approach, a qualitative research strategy was selected. It was determined to use a generic purposive sampling in which ten participants were chosen as part of Generation Z. A semi-structured interview was conducted in which questions were posed regarding the ABC model and the characteristics of AI-powered chatbots. To analyze how attitudes are formed and what they are, grounded theory was used to code the interviews to established consumer attitudes categories. Findings: The findings demonstrate that consumers' attitudes towards AI-powered chatbots used in e-commerce varied from being positive, negative, or mixed. This is due to the factors that contributed to the formation of attitudes.

INTRODUCTION: Conversational agents are of great interest in the field of mental health, often in the news these days as a solution to the problem of a limited number of clinicians per patient. Until very recently, little research was actually done in patients with mental health conditions, but rather, only in healthy controls. Little is actually known if those with mental health conditions would want to use conversational agents, and how comfortable they might feel hearing results they would normally hear from a clinician, instead from a chatbot. OBJECTIVES: We asked patients with mental health conditions to ask a chatbot to read a results document to them and tell us how they found the experience. To our knowledge, this is one of the earliest studies to consider actual patient perspectives on conversational agents for mental health, and would inform whether this avenue of research is worth pursuing in the future. Our specific aims are to first and foremost determine the usability of such conversational agent tools, second, to determine their likely adoption among individuals with mental health disorders, and third, to determine whether those using them would grow a sense of artificial trust with the agent. METHODS: We designed and implemented a conversational agent specific to mental health tracking along with a supporting scale able to measure its efficacy in the selected domains of Adoption, Usability, and Trust. These specific domains were selected based on the phases of interaction during a conversation that patients would have with a conversational agent and adapted for simplicity in measurement. Patients were briefly introduced to the technology, our particular conversational agent, and a demo, before using it themselves and taking the survey with the supporting scale thereafter. RESULTS: With a mean score of 3.27 and standard deviation of 0.99 in the Adoption domain, we see that subjects typically felt less than content with adoption but believe that the conversational agent could become commonplace without complicated technical hurdles. With a mean score of 3.4 and standard deviation of 0.93 in the Usability domain, we see that subjects tended to feel more content with the usability of the conversational agent. With a mean score of 2.65 and standard deviation of 0.95 in the Trust domain, we see that subjects felt least content with trusting the conversational agent. CONCLUSIONS: In summary, though conversational agents are now readily accessible and relatively easy to use, we see there is a bridge to be crossed before patients are willing to trust a conversational agent over speaking directly with a clinician in mental health settings. With increased attention, clinic adoption, and patient experience, however, we feel that conversational agents could be readily adopted for simple or routine tasks and requesting information that would otherwise require time, cost, and effort to acquire. The field is still young, however, and with advances in digital technologies and artificial intelligence, capturing the essence of natural language conversation could transform this currently simple tool with limited use-cases into a powerful one for the digital clinician.

Background: E-commerce is widespread in today’s shopping routines and conversational commerce (CC) as an expansion, aims at integrating customers and businesses on a whole new level. Through the application of chatbots fueled by artificial intelligence, a more personal and individual way of remote shopping is offered. Purpose: Our research question What potential attributes of AI-fueled CC applications and their possible inherent characteristics are determining the willingness to use them and to what extent, in the context of digital natives living in Sweden and Germany? aims at identifying the demanded attributes of conversational commerce from a consumer perspective. Method: We facilitate a quantitative questionnaire with 118 valid answers to administer a traditional full-profile conjoint analysis. Conclusion: Our results indicate that German digital natives deem a CC application’s behavior as the most important attribute, followed by payment method, personality and communication form (voice or text). The Swedish digital natives however, attach the most importance to the payment method, followed by behavior, communication form and personality. Both have in common that they prefer a rather passive behavior over being actively approached, a personality that is balanced between humor and seriousness and text-based communication over voice. A difference is the Swedish preference for direct in-app payment while German digital natives would select a redirection to a secondary payment provider (e.g. PayPal).

It has previously been shown that the quality of the onboarding process affects the chances of a new hire staying at a company, yet it is common that companies have problems in succeeding to maintain a well-organized onboarding. To aid new hires in their onboarding, and to lower the amount of work for the HR personnel, chatbots can be used. In this project, a chatbot was developed for the onboarding process of a large company. Interviews were held with new hires and HR personnel which were used to create requirements for the chatbot. These were divided into two categories; information and functional requirements. A third category, non-functional requirements, was created based on Microsoft’s guidelines for development of conversational AI. Based on the requirements, a chatbot prototype was built using Microsoft Bot Framework with the use of two cognitive services, LUIS and QnA Maker. Both the requirements and the prototype were created iteratively. The information that the interview participants requested from the chatbot was eitherpractical or personal information, or information about the employer, the internal systems, or other employees. It was revealed that the chatbot mainly needed to answer simple questions and didn’t demand any procedural conversation flow which made the use of QnA Maker appropriate. However, for questions and tasks that would benefit or require a procedural flow QnA Maker’s follow-up prompts could be used in future work to create multi-turn conversations.

Detta är en kvalitativ intervjustudie med syfte att undersöka hur supportpersonal upplever chatbotar ur en effektivitetssynpunkt. Fem testdeltagare med erfarenhet av supporthantering intervjuades i studien. Före intervju fick deltagarna genomföra testscenarion med två chatbotar. Testerna genomfördes med data ifrån verkliga och ofta återkommande supportärenden. Därefter intervjuades testdeltagarna om deras upplevelse av effektivisering vid supporthantering med en chatbot. Följande faktorer belystes: tid, kostnad, produktivitet, effektivitet och implementation. Intervjudata antyder att faktorerna är av betydelse för hur supportpersonalen upplever chatbotanvändning ur en effektivitetssynpunkt. Supportpersonalen upplevde att chatbotanvändning skulle kunna medföra både en tidsbesparing och en kostnadseffektivisering. Vidare indikerar intervjudata att en chatbot kan öka supportpersonalens produktivitet liksom effektivitet i supportavdelningen. Resultatet från intervjuerna antyder att implementeringen av en chatbot kan ha en betydelse för såväl tid som kostnad, produktivitet och effektivitet. Detta innebär att hur en implementering genomförs i en verksamhet kan få en inverkan på den sammantagna upplevelsen av effektiviseringen vid chatbotanvändning.

Chatbots are computer programs with the capability to lead a conversation with a human user. When a chatbot is unable to match a user’s utterance to any predefined answer, it will use a fallback intent; a generic response that does not contribute to the conversation in any meaningful way. This report aims to investigate if a content-based recommender system could provide support to a chatbot agent in case of these fallback experiences. Content-based recommender systems use content to filter, prioritize and deliver relevant information to users. Their purpose is to search through a large amount of content and predict recommendations based on user requirements. The recommender system developed in this project consists of four components: a web spider, a Bag-of-words model, a graph database, and the GraphQL API. The anticipation was to capture web page articles and rank them with a numeric scoring to figure out which articles that make for the best recommendation concerning given subjects. The chatbot agent could then use these recommended articles to provide the user with value and help instead of a generic response. After the evaluation, it was found that the recommender system in principle fulfilled all requirements, but that the scoring algorithm used could achieve significant improvements in its recommendations if a more advanced algorithm would be implemented. The scoring algorithm used in this project is based on word count, which lacks taking the context of the dialogue between the user and the agent into consideration, among other things.

The purpose of this study is to form an understanding for communication managers acceptance of the chatbot. This understanding could be useful if the usage of the chatbot at various companies increases. This study also aims to study the relevance of Davis (1989) Technology Acceptance Model, which is one of the most frequent used theories in acceptance studies.  Hence, this study is based on the following questions: <ul type="disc">How can communication managers' perception of the chatbot be understood? Is the Technology Acceptance Model (TAM) a useful model for understanding the acceptance level of the chatbot? If not; how can it be developed to achieve this purpose? The study is based on qualitative interviews of four communication managers from different companies in the Stockholm area. The empirical data collected from the interviews has been analyzed according to thematic analysis, and further understood with support of the TAM-model. Due to the qualitative approach, but also according to the purpose of possibly developing the TAM-model, the study has been characterized by openness and ambition to explore the empirical material and theory. The study showed a unified understanding of the chatbot as helpful in simple types of customer communication areas. The technology was also considered to have a future usefulness in this area. The study also came to clarify that the respondents have shared understandings of the chatbot. The views consisted of both advantages and disadvantages of implementing a chatbot to the workplace. The opinions are based on three emerging themes that were considered relevant to the respondents; the efficient business, future vision for the chatbot and secondary view of the chatbot. The TAM-model proved helpful in understanding the empirical evidence and understanding the grounds to acceptance. The study found that the model would be made even more relevant if it would include social influence, in order to understand how the respondent is expected to understand the users experience of the chatbot. Therefore, this study has proposed an extension of the TAM-model that includes the expected user experience. By developing this parameter to the TAM-model the theory is expected to be more rewarding for further similar studies.

Den nuvarande utvecklingen inom artificiell intelligens kommer att påverka hela arbetsmarknaden, likaså HR-arbetet. Denna kvalitativa studie syftar till att undersöka HR-medarbetares uppfattningar om utvecklingen. Hur kommer artificiell intelligens att kunna användas inom HR och vilka möjligheter och utmaningar för det med sig? Sju semistrukturerade intervjuer med HR-medarbetare inom olika branscher, med olika roller och med viss förförståelse för begreppet artificiell intelligens, har legat till grund för resultatet i studien. Artificiell intelligens kommer enligt informanterna att kunna nyttjas inom de flesta områden inom HR. I dagsläget är det mest aktuellt inom rekrytering, HR-administration och för att besvara HRfrågor av enklare karaktär. Detta öppnar upp möjligheter för mer tids- och kostnadseffektivt HR-arbete, starkare arbetsgivarvarumärke och högre kvalitet på arbetet. Samtidigt möts utvecklingen av utmaningar gällande etik, kompetens och kunskapsbrist samt det mänskliga värdet. Gällande HR-funktionens värde i organisationen finns både chans till ett ökat värde och risk för ett minskat värde. HRmedarbetare kommer i framtiden att tillåtas arbeta mer strategiskt och med mer komplexa frågor till följd av minskat manuellt arbete, något som kommer att gynna organisationen på olika sätt i längden. Om robotar ersätter den mänskliga arbetskraften riskerar HR-funktionen dock att tappa sitt värde, däremot kommer nya arbetsuppgifter som efterfrågar HR-kompetens växa fram. Trots utmaningarna ses utvecklingen som spännande och är något som HR-medarbetarna är nyfikna på och ser fram emot.

La gestión de información de la calidad en las universidades del Perú es uno de los factores más importantes para el desarrollo académico; sin embargo, no todas las entidades de educación superior cumplen con los estándares de calidad y políticas. Además, el acceso a toda la información de la calidad es muy difícil de administrar, si la universidad cuenta con un sistema muy limitado, sin mantenimiento y actualizaciones. La Universidad Peruana de Ciencias Aplicadas posee el Sistema Integrado de la Calidad de UPC, en adelante SICA, en el cual se publican los diferentes documentos oficiales de la universidad, tales como reglamentos, políticas, manuales, rúbricas, entre otros. Sin embargo, el proceso de búsqueda del portal web es muy limitado y engorroso, lo que conlleva a experimentar dificultades en la búsqueda de información por parte de los usuarios. En este proyecto se propone un sistema cognitivo para mejorar la experiencia del usuario en la búsqueda de información académica con un chatbot. La diferencia entre los sistemas de búsqueda tradicionales y el sistema cognitivo propuesto es mejorar la experiencia del usuario (UX) a través de factores de optimización como el tiempo de respuesta, la facilidad de uso, la interfaz amigable y la interacción del usuario a través de los servicios cognitivos de computación en la nube. Esto puede ir más allá de la interacción entre un chatbot y un humano, ya que la experiencia del usuario es muy importante y puede definir el éxito o el fracaso de un sistema. / The management of quality information in the universities of Peru is one of the most important factors for academic development; however, not all higher education entities comply with quality standards and policies. In addition, access to all quality information is very difficult to administer, if the university has a very limited system, without maintenance and updates. The Peruvian University of Applied Sciences has the Integrated Quality System of UPC, hereinafter SICA, in which the different official documents of the university are publisher, such as regulations, policies, manuals, rubrics, among others. However, the process of searching the web portal is very limited and cumbersome, which leads to experiencing difficulties in the search of information by users. In this project, we propose a cognitive system to improve the user experience in searching for academic information with a chatbot. The difference between traditional search systems and the proposed cognitive system is to improve the user experience (UX) through optimization factors such as response time, ease of use, friendly interface, and user interaction through cognitive services cloud computing. This can go beyond the interaction between a chatbot and human since the user experience is very important and can define the success or failure of a system. / Tesis