Embedded Files
When I read the post of Katerina Saiti on uncertainty, I found this quality of creatureness very relatable as a human and interesting to explore in artificial creatures, especially by the means of fortune-telling as done in her post.
To expand on the concept of the Oraculon (an automata that can tell your future) I developed a potential design for the Oraculon, along with several additions and improvements.
Before creating the design, I conducted a brainstorming session on Katerina Saiti’s ideas for the Oraculon. I sketched various elements that she described in her post, to understand how the design should work and look.
Important in Katerina Saiti’s blogpost, is the Barnum effect. According to this effect, vague or inaccurate fortune-telling can gain personal meaning through interpretation by the spectator. I was wondering if it would be possible to make the told fortune by the Oraculon more accurate, instead of only making use of this effect (which is essentially a con-technique).
I thought of an article by the New Scientist about AI algorithms that are predicting the future. In the article, the AI predicted the future based on extensive historical data. It could predict the future for five years with more than 99% accuracy. Could we use AI to predict our own, personal future? Big tech companies such as Google and Facebook already have personalised advertisement profiles that predict what we want and will buy entirely based on our digital data.
To make the Oraculon more interactive and personal, I want to use the spectator’s personal data in a similar way. To collect this data for a prediction, the spectator can connect their smartphone to the Oraculon. The told prediction will be personalised based on the personal data found on the smartphone and told as a cryptic fortune through an AI in the form of a generative language algorithm.
To illustrate how the AI could work, I tested fortune-telling prompts in AI chat GPT. I used recent WhatsApp messages as my personal phone data.
For the next step, I developed the interaction design of the Oraculon. This is done in the form of user tasks:
· Start
o Press start button
o Music plays
o Curtains open
· Instructions
o Oraculon speaks:
o I can tell your future
o Connect your phone
· Collect personal data
o Connect phone with USB cable
o Data transfer to Oraculon
· Fortune telling
o Personal prompt for generative language model is created with the data
o Generate fortune
o Share fortune (speaking and print)
· End
Once I had a clear understanding of the interaction, I created the full physical design of the Oraculon.
I think the design for Oraculon can address uncertainty and make us think about our future, while simultaneously making us aware of how sharing our data can influence that future.
Read more
Katerina Saiti’s post: https://sites.google.com/view/artificial-creatures-2023/portfolios/katerina-saiti/oraculon?authuser=0
New Scientist article about AI predicting the future: https://www.newscientist.com/article/2342632-artificial-intelligence-is-being-asked-to-predict-the-future-of-ai/
Page updated
Google Sites
Report abuse