Discussion

The success of interaction design depends on human psychological underpinnings. In the following tables I’ve used the Theory of Planned Behavior to categorize each feature according to what psychological state they aim to affect. To have any chance to impact human behavior towards sustainability, the nudges should be delivered in-context, and at the right time, using effective language. This section will focus on key findings that could be relevant for designing the right context - i.e. inside particular Features of the experience.

Design Implications

The following findings from Literature Review, User Survey, and Prototype Testing, have been categorized by chapter and list some of the key implications which can lead to Feature Ideas.

College

College students need tools for action. General theory about Taiwanese society suggests that it’s low-context and people need groups to do activities. But my survey about my sustainability app shows that people don’t pick the features for joining groups. Joining groups was one of the least interesting choices in the survey.

TPB	Design Factors	Potential Explanation for Group-Related Features
Attitude	Survey Framing Effect	Survey respondents may feel the way I framed the survey question for “joining groups” made it sound like an extra commitment they didn’t want to make.
Perceived Control	Group Fatigue	Survey respondents may already have too many group obligations to attend to (school, work, family, etc). Joining another new group may feel like a burden, not a benefit.
Perceived Control	Fear of Awkwardness	Survey respondents may feel joining strangers in online group is risky, uncomfortable or unnecessary.
Attitude	Desire for Personal Achievement; Changing Cultural Trends	Survey respondents may be more motivated by individual achievement to personally feel they are making a difference instead of waiting for the group. Possibly they are more individualistic than my expectations in their behavior (particularly in a digital context) and the traditional “group society” stereotype is weakening.
Subjective Norms	Are “Group” and “Sustainability” Concepts Related?	Survey respondents may not clearly understand the benefit for sustainability from their joining a group.

Environment shapes action.. how to create an environment where college students can influence companies.

Design implications arising from the college chapter.

TPB	Human Interface	Findings and Roles	Design Implications
Subjective Norms	Community	Taiwanese students are highly influenced by the actions of their peers. People exist in relation to other people.	The app should show what other people are doing. In terms of a specific Feature, this could become “Group Purchases”, “Find Your Composting Community”, “Create a Group Chatroom”.
Perceived Control	Investing	Psychology of ‘fundraising clubs’ vs individual investing differs greatly.	Provide a community for pooling money with like-minded investors.

Sustainability

Design implications arising from the sustainability chapter.

Design implications arising from the sustainability chapter.

TPB	Human Interface	Findings and Roles	Design Implications
Attitude	Shopping / Transparency	Realtime ESG is a building block to enable consumers and investors make more accurate, real-world purchase decisions.	Build technologies and practices that enable real-time ESG: up-to-date transparent information about how our products are produced.
Perceived Control	Shopping / Actionability	Start with the most polluted regions as priority?	Simplify action towards a “triple turn”: environmental, social, governance.
Perceived Control	Shopping / Pollution	People living in polluted areas are so used to it.	What app to wake them up? “You live in a highly polluted area. Here’s the TOP 10 companies causing pollution. Here’s what you can do.”
Perceived Control	Saving / Health Tracking	Blood testing and biomarkers allow people to track their health.	I’m introducing the concept of ‘eco-markers’ to follow the sustainability of human activities.
Attitude	Circular Economy	AI can help us make sense of the vast amounts of sustainability data generated daily.	Sustainability is part of product quality. If a product is hurting the environment, it’s a low quality product.
Perceived Control	EPR	There is a lot of new legislation (especially in Europe) encouraging sustainable design.	EPR and CDP data should be part of Green Filter.
Perceived Control	Eco-Design	As a consumer, eco-designed products are hard to find.	Provide clear labeling to find eco-designed products.
Subjective Norms	Governance	Politics matter and lack of transparency is intentional.	Make it easy to support policies that call for transparency.

Design

This chapter has the following design implications.

Several superapps already contain features for payments (shopping), savings, and investing. Yet none thus far integrate Digital Product Passports to understand the products’ journey, including the origin and manufacturing conditions, materials, components, CO2e footprint and environmental impact, post-consumer repair, recycling, and end-of-life disposal guidelines.

Design implications arising from the design chapter.

TPB	Human Interface	Findings and Role	Design Implications
Perceived Control	User Interface	AI now enables generative UIs, which can dynamically change the layout and content as needed and fitting to the data.	It’s a balance: while there’s potential, users also need some type of stability (think: text input stays in the same place but different types of interfaces appear within a clearly defined space).
Attitude	Health	Many respondents express concerns about health.	Show health-related alerts in addition to sustainability. For example, Aspartame has not been banned or reduced after 2 years of findings; provide warnings for possible cancer-causing ingredients in the AI assistant.
Subjective Norms	Greenwashing	Personal CO2eq tracking is ineffective and the focus should be on systematic change towards circular design and zero waste practices.	Provide tools to influence companies instead of only personal lifestyle changes.
Perceived Control	Goal-setting	Users feel bored if there’s no updated information.	As a student, I can set an Intention for myself, such as cutting plastic waste or building a climate positive investment portfolio.
Perceived Control	Accountability	Consumers lack tools to affect change.	Help consumers to demand more by integrating tooling into the shopping experience. Perhaps this could be called “embedded sustainability”, inspired by the “embedded finance” trend documented in literature.

AI

This chapter looked at AI in general since its early history and then focused on AI assistants in particular.

Design implications arising from the AI chapter.

TPB	Human Interface	Findings and Roles	Design Implications
Perceived Control	Ambient Computing	AI companions could combine sensor data from human bodies with the ability to reason about human speech, to provide increasingly relevant, in-context assistance.	Because of the conversational nature of LLMs, they are very useful for affective computing. OpenAI is developing such a device.
Perceived Control	Voice Assistants	There are many distinct ways how an algorithm can communicate with a human.	From a simple search box such as Google’s to chatbots, voices, avatars, videos, to full physical manifestation, there are interfaces to make it easier for the human to communicate with a machine.
Attitude	Sustainability	While I’m supportive of the idea of using AI assistants to highlight more sustainable choices, I’m critical of the tendency of the above examples to shift full environmental responsibility to the consumer.	Sustainability is a complex interaction, where the producers’ conduct can be measured and businesses can bear responsibility for their processes, even if there’s market demand for polluting products.
Attitude	Sustainability	Personal sustainability projects haven’t so far achieved widespread adoption, making the endeavor to influence human behaviors towards sustainability with just an app – like it’s commonplace for health and sports activity trackers such as Strava – seem unlikely.	Personal notifications and chat messages are not enough unless they provide the right motivation. Could visualizing a connection to a larger system, showing the impact of the eco-friendly actions taken by the user, provide a meaningful motivation to the user, and a strong signal to the businesses?
Attitude	Cuteness	Cuter apps have higher retention. Literature suggests adding an avatar to the AI design may be worthwhile.	Design the app to be cute and ask the user for their favorite animal.
Subjective Norms	Anthropomorphism	People lose trust in AI if it lies.	AIs should disclose they are AIs. Understanding algorithm transparency helps humans to regard the AI as a machine rather than a human.

Finance

Design implications results from the literature review on finance.

TPB	Human Interface	Finding and Role	Design Implications
Perceived Control	Legislation	As a consumer, legislation does not always protect me from being complicit in pollution, even if unintended.	As a consumer, I can get notified by the app about highlights of poor legislation refuted by science.
Attitude	Greenwashing	As a consumer, sustainability is fragmented and greenwashing is widespread: how can I feel trust, honesty, and transparency?	As a consumer, I can make use of Green Filter, a sustainable shopping, saving, and investing companion.
Attitude	Greenwashing	As a consumer, while reading the EU Commission’s proposals shared in the news, I might think the politicians have everything under control, I can relax and continue the same lifestyles as before, the reality is emissions keep rising, while they should be falling.	As a consumer, I need proper tools to understand what’s sustainable. I want to know it’s possible to curb greenwashing.
Attitude	Education	As an interaction design student who cares about the environment, I ask myself how can interaction design contribute to increase sustainability?	I make the assumption that investing is inherently “good” for one’s life, in the same way, that doing sports is good, or eating healthy is good. It’s one of the human activities that is required for an improved quality of life as we age. To start investing sooner, rather than later, because of compound interest. Nonetheless, investment also includes higher risk than sports or food.
Attitude	Education	As a designer for a financial product, I need to communicate the risk of investing effectively while educating the users.	?
Perceived Control	Metrics	As an investor, I want to know where my money is going.	As an investor, I can visualize what happens with the money.
Perceived Control	Metrics	ESG can’t be trusted. ESG-reporting alone is not a sufficient metric to prove sustainability of a company.	ESG needs to be accompanied by other metrics.
Subjective Norms	Governance	As a consumer or investor, I can’t trust the people running the company.	Show company board membership in the app product view.

Research Limitations

First, this work is focused on integrating sustainability into user experience design and does attempt to make a contribution to economics, finance, on any related technical field - in the context of this work, these fields are sources of inspiration for design.

Second, finance is a highly regulated industry and the proposed user experience designs may be limited by legal requirements. This study does not take such limitations into account, rather focusing only on the user experience.

Third, up to date ESG data is expensive and thus couldn’t directly be used in this research. ESG needs data to give us an accurate understanding of the realities inside companies and the user experience design does not address the underlying data quality problem further than by providing a link to the data source.

Fourth, I don’t have access to users’ financial data, which would be useful for design research.

Future Research

Some ideas for fruitful research directions:

• Perceptions of pollution levels among Taiwanese college students.

• Does the specialized interface for AI offer any advantages of a general-purpose UIs such as ChatGPT, Claude, Gemini, Mistral, and others?

• While many people are working on AI models, there’s a lack of people working on Human-AI interaction in the context of sustainability.

If you do decide to pursue any of these questions or were otherwise inspired by my thesis, please do reach out. As I have interest in these areas of research, I would happy to help in any way I can. Thank you.