Title

Federica Santoro - Università di Bologna

创造一把中式紫砂茶壶，壶的形态不限。壶身有唐诗，壶身有中国神仙的元素，壶身有竹子的元素。这把茶壶的主要用户是中国当代的年轻人，设计需要让年轻人容易接受传统茶壶。 

Persona profile

Feedbacks

Great description! Very good concept!

Your creation of the user profile is very successful. The name, background, working environment, etc., fit the expected image of a contemporary young girl in China.

You have a good understanding of the usage and interest points of traditional Chinese teapots, especially in the aspect of the design concept, you mentioned the multifunctional use method of teapots, which is a good design concept that makes Chinese contemporary young people accept traditional teapots more easily. It's also a cool idea to use sculptural decorations to show the amount of water.

It should be noted that there are many schools of traditional Chinese teapots. Therefore, in my brief, I particularly emphasized the teapots of the "Zisha (Purple clay) teapots" school, which require special Zisha minerals to make and the firing method is different from that of ceramics. If you can highlight this feature in your design, it would be perfect.

Concept

Feedbacks

leave feedback here

Notes and reflections

Hunting report

Brief for Zheuri

The brief chosen for Zheuri consists of a 'sponzafrise', a specially shaped dish for eating a typical food, friselle. Preliminary tests were conducted on ChatGPT in Italian and English for an initial evaluation prior to the final choice. 

The sponzafrise is an existing but uncommon object, typical of the cultural context of reference. Morphologically, it is a bowl, and therefore a concave-shaped object, containing water used to dip the friselle. The bowl is partially covered by a lid with holes to drain excess water. The finish usually recalls the local ceramic tradition through colours and decorations.

Brief from Zheuri and persona

The brief assigned by Zheuri consisted of a particular teapot from the Chinese tradition, needing a reinterpretation for a young target group. The goal was therefore to propose a new shape that could combine tradition and contemporaneity, culturally well-accepted for the young Chinese population.

In the first phase, two ethnographic researches were conducted in parallel through ChatGPT. The first concerned the Chinese tradition, of the teapot Zisha specifically exploring most commonly used materials, typical shapes and decorations of the Chinese tradition. This in order to obtain an overall background on the cultural elements that defined the canonical form of Zisha teapot. 

The second research aimed at the construction of the personas and gaining knowledge around tea rituals in China. Again, the main focus was aligned to the target audience emerging in the assigned brief. I asked ChatGPT for an initial personas profile corresponding to a young person of Chinese nationality, asking for more specific information from time to time in order to better delineate the profile to an appropriate completeness. The image generated by Midjourney was generated from information obtained with ChatGPT about the person in question: a young worker in Shanghai.

The information generated was broadly generic, subtle bias emerged but the feedback from my mate was found to be plausible in the end. 

Chinese teapot - design concept

For creating the design concept I started by entering a summary of the brief, key-features of persona and then asking the tool to formulate an array of How Might We and project requirements taking into account the user's lifestyle and difficulties. 

These were then transposed into a teapot concept, asking each time to refine the result. Traditional materials and shapes were a strict constraint. Soon, the possibility of creating a multifunctional object that could help in the preparation of tea emerged. 

ChatGPT's decorative proposals tended to reformulate what contained in the initial brief, not coherently integrating information. When too far from the constraints set within the brief, attempts were made to realign the research to the initial goals and constraints.

In the concept refinement phase, an attempt was made to explore both functional and social aspects of the proposed HMW with ChatGPT. The tool generated several design ideas for the former, the functional aspects. The social insights are difficult to gain, responses generated had poor definition and not well-integrated with the object itself and not bringing out insights into the social value of the tea ritual in China.

Sponzafrise - personas

The generated person may be plausible in terms of appearance even if bias emerged.

The persona profile regards an Apulian man, any information about his occupation and lifestyle insist in tourism-related activities in the Apulian region, with little to no variety. Also his first occupation in Northern Italy might have been generated from stereotypes regarding patterns of migration from south to north in Italy.

There are some common somatic traits generally associated with southern Italy: dark complexion, curly hair. The technical-sports clothing, and in particular the orange backpack, suggests that the result was influenced by the amount of words related to tourism.

The houses on the scenery may be the result of stereotypes about small towns, the buildings are low with uneven plaster. 

The main error of the tool emerges in the representation of the food and related eating gestures: what may be more generally a bruschetta with a saucer is depicted, which is an object unrelated to the consumption of friselle, as is the seasoning - resembling some kind of cheese - depicted.

Sponzafrise - design concept

Zheuri's output consists of a plate, so ChatGPT has broadly identified the correct product sector. 

Midjourney correctly returns the concave shape of the object, but leaves out any potentially useful component intended for draining excess water or for holding food seasonings. It must also be taken into account that, referring to the image of the personas, the term 'frisella' was not interpreted correctly in the first place, generating instead in the image what is more similar to a bruschetta.

In the image generated by Midjourney, it is presented a light-coloured bowl with decorations on the rim. Comparing it with typical Apulian ceramics, similarities can be recognized within the pattern, its position on the product, and concentric blue lines on the rim. The decorative lines are a direct match; the decorative motifs generated with Midjourney appear to be olives, whereas stylised flowers and more generally flat colours are common in the tradition. 

However, I think that this result may have been influenced by the association of 'olive' with 'puglia' or the presence of the word 'olive oil' in the given prompt.

Although not directly recognisable, analysing the process and the limitations of the tool, it is a fairly satisfactory and consistent result speaking of the decorative pattern. 

Conclusion

In conclusion, the main difficulties arising from the use of artificial intelligence concern the lack of direct observation of a human person. ChatGPT responses often proved to be repetitive and inconsistent, formulated and recombined on the basis of prompt words and common associations. 

The social and contextual elements even when vague or subtle are fundamentally based on the elaboration of stereotypes emerging particularly in the places, lifestyle, occupation of the personas. The limitation of the tool is that it does not allow an in-depth ethnographic analysis to be conducted because it quickly reaches redundancy of information.

Artificial intelligence, as demonstrated in both projects, mine and my mate's, returned demonstration of quite good morphological knowledge of products and was able to respond to functional problems with different design insights. However, was almost completely lacking a critical analysis of the human data and response to the person's gestures for morphological elaboration.