Skip to main content
Login | Suomeksi | På svenska | In English

Browsing by Subject "Generative AI"

Sort by: Order: Results:

  • Investigating Motivation and Usage of Text-to-Image Generative AI for Creative Practitioner 
    Anyatasia, Fayya (2023)
    This study researched the rationale behind creative practitioners for utilising or not utilising Text-to-Image Generative (TTIG) AI in their creative process. In addition, it also researches how the workflow of creative practitioners who are utilising this technology. A Reddit analysis of 331 posts and their comments and an online questionnaire with 92 participants is performed. The result showed that the rationale for creative practitioners not using TTIG is varied, including personal reasons, impact on the artist, ethical issues surrounding it, the AI-generated art itself, and their own creative workflow. On the other hand, motivation for using TTIG is mostly driven by the playfulness and usefulness of the system. Amplified by the benefit felt by the users for example source of inspiration, helping idea generation and exploration of new creative possibilities. There are mainly four types of workflow incorporating TTIG: to use it as reference only, use it as is, use it as a base, and use it as parts. We further discuss the implications of these findings and the author highlights the urgency of policymakers to create regulations safeguarding creative and their creations. The author also proposes to develop the system collaboratively with creative practitioners and the inclusion of AI in art education curricula.
  • Utilizing Generative Artificial Intelligence for Tax and Legal Consultancy: Design Science Approach 
    Roberts, Taylor (2024)
    The continuous application of generative AI is increasingly employed to enhance both creative and knowledge-centric processes. With the release of OpenAI’s GPT models in 2022, generative AI utilisation has surged, facilitating the optimisation of these processes to a greater extent than before. This research has been conducted for PwC Finland and presents a solution that leverages generative AI and RAG methods to aid consultants in data retrieval and analyses. This work follows a design science methodology, whereby, an incremental and iterative process is followed to produce an IT artifact. The artifact is in the form of a Proof of Concept (POC) instantiation: a Microsoft Word Add-in that links to a backend process employing a RAG framework, generative AI, and contemporary software architecture. This research has found that utilising RAG methods with generative AI alone is not enough to produce specific analyses within the tax and legal context. With an accuracy of only 62.5%, its important to utilise metadata filtering, pre-prompting and knowledge graphs to enhance contextual understanding.

Yhteystiedot

HELSINGIN YLIOPISTO