Artificial Intelligence-powered Data Thinking

^{Florian Straetmanns, Furkan Dursun} 

Dialogue-based generative AI models, such as ChatGPT and Google Bard, have democratized the use of AI and made it accessible to the general public [1]. As a technological revolution, they offer a wide range of applications and benefits, including in creative fields where they can support humans in various ways [2]. In view of the public response and the rapid spread of ChatGPT since November 2022, as evidenced by over one million user profiles, it can be assumed that this technology is here to stay and will continue to play a prominent role in various scenarios in the future. Its benefits will be realized in particular in the form of automation and collaboration between humans and AI [1].

As part of our AID Thinking project, we will investigate generative AI (genAI) in particular in relation to creative processes and the idea generation process. Before the emergence of generative AI models, studies often considered idea generation and prototyping in the innovation process as less important areas for AI applications [3]. However, this view has changed [4-6]. The new capabilities of genAI to generate content - a feature that was not foreseeable in earlier studies - have the potential to significantly influence various industries [7]. Even human activities that were previously considered impossible to automate, particularly in the area of text generation, are now being called into question [8].

In the context of this objective, the AID Thinking project aims to evaluate the extent to which genKI is applicable in the context of idea generation and evaluation based on current scientific literature and practical findings. In particular, the aim is to identify the advantages and disadvantages as well as the limitations and potential of this technology. One focus is on its possible role in supporting thought processes, structuring ideas and objectively evaluating them. The focus is on the conception, development and testing of a format for idea generation and evaluation supported by ChatGPT for data thinking, supplemented by the consideration of functionally extended plugins. On this basis, application principles for the use of the developed genKI format in the innovation process will be derived. The focus here is on aspects such as the integration of plugins, the targeted control of the dialog process and the effective communication of approaches. These principles are to be developed taking into account current technological limitations in order to ensure the targeted use of genKI. The overarching goal is to be able to adapt the AID Thinking format to different contexts and thus enable versatile applicability. The results can help stakeholders to develop a better understanding of the use of genKI in innovation processes.

To investigate and address these aspects, the AID Thinking project is based on a data-driven approach. It starts with a literature review to capture the current state of the art. In the second step, current plugins of ChatGPT are examined. These plugins are assessed for their applicability in the context of idea generation and evaluation. If there is potential, these plugins will be included in step 3. Based on the data from steps 1 and 2, we will develop the AID Thinking format in step 3, which is currently planned to be based on ChatGPT. In step 4, the developed AID Thinking format is tested and revised according to the design-oriented Design Science Research paradigm [9]. At least two trials are planned, each with 25 or more participants. Steps 3 and 4 are iterative, i.e. after the first trial, the format is revised and further developed based on the findings before it is run through a second time. At the end of the runs, surveys will be conducted with the participants to record their feedback on human-AI collaboration and to quantify the perceived benefits of genKI. In addition, during the trial it is planned to have at least one group act without genKI support in order to compare the differences in the results. In the final step 5, application principles for the use and adaptation of the AID Thinking format will be established based on all the data collected.

And here the group in their own words (German):

References

[1] Bilgram V. and Laarmann F., (2023). Accelerating Innovation with Generative AI: AI-augmented Digital Prototyping and Innovation Methods, in IEEE Engineering Management Review, https://doi.org/ 10.1109/EMR.2023.3272799.  
[2] Larsen, B., and Narayan, J. (2023). Generative AI: a game-changer that society and industry need to be ready for. World Economic Forum. Retrieved from: https://www.weforum.org/agenda/2023/01/davos23- generative-ai-a-game-changer-industries-andsociety-code-developers/.  
[3] Füller, J., Hutter, K., Wahl, J., Bilgram, V., Tekic, Z., (2022). How AI revolutionizes innovation management – Perceptions and implementation preferences of AI-based innovators, Technological Forecasting and Social Change, vol. 178, 2022, https://doi.org/10.1016/j.techfore.2022.121598.  
[4] Bilgram, V., Canadas Link, D., Lang-Koetz, C., (2023). Generative KIs in Kreativprozessen: Praxiserfahrungen aus den ersten Monaten mit ChatGPT & Co., Ideen- und Innovationsmanagement, vol. 1, pp. 18-22.  
[5] Brem, A., Giones, F., Werle, M., (2023). The AI digital revolution in innovation: A conceptual framework of artificial intelligence technologies for the management of innovation, IEEE Transactions of Engineering Management, vol. 70, no. 2, https://doi.org/10.1109/TEM.2021.3109983.  
[6] Bouschery, S., Blazevic, V., Piller, F., (2023). Augmenting human innovation teams with artificial intelligence: Exploring transformer-based language models, Journal of Product Innovation Management, vol. 40, https://doi.org/10.1111/jpim.12656.  
[7] Agrawal, A., Gans, J., & Goldfarb, A. (2022). ChatGPT and how AI disrupts Industries. Harvard Business Review, pp. 1-6.  
[8] Burger, B., Kanbach, D. K., Kraus, S., Breier, M., & Corvello, V. (2023). On the use of AI-based tools like ChatGPT to support management research. European Journal of Innovation Management, 26(7), 233-241.  
[9] Hevner, A., March, S. T., Park, J., Ram, S., (2004). Design Science in Information Systems Research. MIS Quarterly, 28, 75-105, https://doi.org/10.2307/25148625.