Questions and Answers
Hi to everyone, just to trigger the discussion. Do you think Generative AI is helpful in data storytelling?
I think definitely yes! Generative AI helps people who are not good at Data storytelling, coding or data visualization.
But it also let user to limit their imagination for creating or establishing their own ideas.
ps. I just start to read this book “ Data Storytelling with Altair and AI”–truly amazing!
I also wonder if anyone use AI to create dashboard-if yes, could you please share the expereinces with us?
I like the idea that AI can increase imagination and creativity if you use it as a companion or friend with whom you can discuss. My boss says that AI means Augmented Intelligence and not Artificial Intelligence. For example, I have used AI to improve a poetry I wrote!
Following the same line, you can use AI not only to create your dashboard but to organize the visualizations based on the audience, or even to ask AI which metrics you should monitor in your case
AI is a wonderful assistant which you can ask everything without shame
Thank you, Angelica! I also would like to know how your company deals with AI mapping/drawing? Because this issue is very polarised in Taiwan, some people think it’s helpful, others think it’s going to put a lot of designers out of a job.
Your question is very interesting. I work in a research institution which uses AI only for research and not for business. However, I think that using AI for drawing is opening also ethical concerns. I have recently read a post which says that in Brazil children photos have been stolen from private blogs and social media to train AI models
Thank you for your reply. And really sorry to hear that the children’s photos have been stolen in Brazil society…
This is a very keen topic of interest for me and I have started reviewing this wonderful book. Thank you Angelica Lo Duca. One thing that I have concern about generativeAI is while its extremely helpful for cases where we use it as an assistant or supportive framework in order to consolidate or even data mining and feeding us with key information to guide us making data decisions. On the other hand during real customer use cases at times raises the question of data quality that is fed into the dashboards. Could I ask if there are different ways to measure the quality of data that is refined by generative AI to continuously measure the quality? More specifically what are some common measures we can use to measure the quality of extracted content from genAI before we use it for key dashboards or decision making.
This point is very interesting. Thanks for the question. Just recently, I’ve been working on a project where I tried to evaluate the results of analysis produced by AI. We are developing a system that is divided into two parts. During the first phase, we calibrated the AI model with some data. Calibration is done manually with the help of a domain expert. When model calibration is completed, we can move on to the production phase, where the model is run on all the data. However, the expert should verify the results on a sample basis. I have attached the designed system here.
I’ll present the associated poster paper at the Discovery Conference 2024 http://ds2024.isti.cnr.it/