GreenWare - Rethinking AI Development with Ethical and Environmental Considerations

Our Team

Our Problem: How can we Address the Sustainability and Environmental Impact of AI Development

• Human-centered around the carbon footprint of new human-developed AI programs
• This is a relatively new topic, and not something that has been extensively studied, making it a complex problem with societal, technological, and environmental impacts
• It has impact on today’s world and future generations, especially given the current state of the environment

Empathize

Research Plan:

• We decided to find potential solutions to our problem by conduct various methods of research
• Case studies and interviews were two primary methods undertaken
• We researched throughout the design thinking process whenever possible
• We primarily explored how AI works and how it can impact the environment

Define

Who are the users?

• In the past AI has been a relatively niche field but as we see an increase of its integration into everyday life, more and more people interact with it
• Everyone is reliant on the environment, whether they are conscious of it or not

We needed to understand their needs and goals - how would we do this?

• We conducted interviews to test the waters of people’s opinion both of AI and of its effect on the environment so that we could create a baseline for our development

Ideate

Finding creative solutions

Once we had a baseline for our research via our interviews, we needed to flesh out our product’s direction

• We knew we were working with limited time so what tools do we have?
• Brainstorming sessions
• How Might We?

How Might We?

We used a brainstorming process called How Might We, where we imagined a subject and then created various “how might we?” questions concerning our research

Ex:

• How might we develop AI programs without the large infrastructure requirements that it takes now?
• How might we develop AI to create ethical solutions to environmental problems?

These questions made us think outside the box when it came to our approach to the problem and helped us create a more well-rounded idea

Personas

One of the ways we kept our design process user-centric was to create personas

• This put us into the position of the user
• It gave us a deeper understanding of our audience
• Forced us to tailor our design to the potential user’s specific needs

The creation of our personas also allowed us to connect as a team because it made us focus on one common individual, their thoughts, needs, wants, etc

Our goal was to flesh out our idea in as short amount of time as we could so we could create a well rounded prototype

Prototype

Forced Connections + Prototype + Voting

• Created two categories, experiences and products, to try and find new, unexpected connections
• Our mashups allowed us to generate concepts for products
• Further set of connections and concept generations followed
• We mapped out our concepts in a matrix to understand the overlap and gaps
• We then evaluated and voted on our main concepts to know what our product could be
• Once we had a product concept selected, we sketched out prototypes to get a better idea of what our product is.

Test

Service Blueprint

• We took our ideas and created service blueprints to go along with them
• These were to allow us to better understand the interactions with users and processes that influence important steps in the creation of a product.

Iterate

Reflecting on prototypes - through storytelling and user testing

• We were able to take our prototypes and narrow them into the one we wanted to bring to users and get feedback.
• We did this by creating a story with our prototype that then with questions we shared with stakeholders.
• Through the feedback and questions, the next step is to evolve our prototype to be the best fit.