An Update
Ideation, creation and the scariness of reaching out.
Happy Midterm season!
Clearly I’ve been celebrating by not writing anything…
All jokes aside, many things have happened since I last updated you all. At this point, most of our project has come to fruition via powerpoint, but I’d be thrilled to walk to through the steps.
I was lucky enough to be paired with two other amazing people who have studied abroad at some point. Our combined experience of dramatically improved mental health (because of the walkability of these places) spurred a long thought process about how we could make more places walkable. From this thought spurred two ideas:
Since the prompt for the project had to do with emerging technologies of some sort, we thought it might be interesting to use AIto help with city planning and making cities more walkable, as well as incorporating more green spaces into said cities. However, we then realized that we could not, in good conscience, use ai to help environmental developments without first assuring that it wouldn’t harm the environment in doing so. (Spoiler alert), of course it’s harming the environment, and after some preliminary research where we learned from the Federation of American Scientists that “Data centers consumed an estimated 415 Terawatt hours (TWh) of electricity in 2024 (roughly 1.5% of global power demand). Lovely! We concluded from our research that it would be ridiculous to use AI in our urban planning endeavors when there were bigger issues at stake.
We learned so much about AI in fact, that we started to wonder how much other people knew about it, and that if they really knew what the cost of using it was, maybe they would use it for less arbitrary things, such as making a grocery list or answering all of their health questions. Thus the question:
“What are people’s perceived environmental impact of AI usage versus the real impact?”
was born.
This lead us to create a how might we statement. Based on our own understanding of how we know about AI, and the people in our immediate circles, we thought that
“How might we interpret people’s understanding of environmental impacts and where they derive their opinions?”
seemed the most fitting. To first understand how to get to people, you need to know where they get their information, how they learn and who thay talk to. After learning about how big AI is and how complicated it would be to solve the genuine engineering/code/whatever problem, it seemed more feasable for us as a group of students to go to the source of its continuing existence, it’s users. Knowing this, we then came up with a job to be done: make strides in lowering the environmental impact of AI.
How you might ask? The hope is to help equip people with information they need to be informed about how AI impacts the environment to encourage responsible use if any. As stated in my previous (ancient) post, I really don’t think that AI will just go away. But if we can learn to use it responsibly, we would not only be helping the environment, but keeping our own brains sharp and independent. There’s nothing wrong with getting help, but when you have someone else do everything for you (be it human or AI), eventually you lose the autonomy to function on your own.
We created empathy maps and personas of people that we imagined would use AI on a daily basis, including tidbits about it that we’ve heard from real people in their hears and thinks sections. We spent time thinking about where exactly we would need to research to understand how to best reach the people who need to know about AI the most. What was particularly interesting to me about this process was how often I saw my experience using the design cycle in high school and middle school was needed. I went to an IB school, and often complained about having to take design courses because they were executed awfully. However, I will say that it did cement into my brain how to break down a problem, and think through what needs to be true to get to our end goal. This is a strong suit that I have seen play out across all fields of work and education I have applied myself to, but particularly in this process, it has flourished. What has seemed obvious to me I don’t think was always as obvious to others, and between their expertise in urban planning and design, and my knowledge of the design cycle and ideation, I think we had a really well balanced and effective team.
Through this process of ideating and talking it out, we came to the conculsion that we needed three main things from our research methodology:
- Understand metrics and units (gigawatt, gallon, kilowatt, etc) used to measure energy usage of AI
- Develop a survey that aims to understand how people perceive environmental impacts
- Conduct intercept interviews on campus and at local coffee shop to gather a variety perspectives with difference experience and educational backgrounds
We realized that many people (including ourselves) really didn’t know the magnitude of the difference, or how big the numbers were because we didn’t even know what we were measuring in. On top of that, it was important to just gage where people were with understanding AI, not just from our friends, but from all sorts of people.
Get Emily Milo’s stories in your inbox
Join Medium for free to get updates from this writer.
To execute this, we decided to create a Qualtrics Survey. The most important thing to us while creating this survey was that it remained neutral on the stance of AI, so that participants would feel comfortable enough to be open and honest about how they interact with it. We started with some open ended questions. General and easy ones like
Open Ended Questions:
How old are you?
What is your area of study or career field?
After that we opened into giving them the option to tell us 3 things they know about AI, with no descriptor word purposefully used for confortability. As well as a space for additional thoughts.
We then asked about where their information about AI came from, and gave them options to choose from, as well as a space to specify, as well as how they think that AI language models are powered, and what AI’s contribution to the environment is. We conducted interviews both on campus and at Wake the Dead Coffee House, which brought all sorts of walks of life. It was my job to go to Wake the Dead, and I actually had a really fun time talking to people. There were conservative and liberal regulars, people already in the workforce and students meeting up to work, professors and grad students. It was kind of scary just walking up to people, but after a little while I got used to it, I even spoke to some of the people that work there. Below is a map of some of the most used words from our interviews:
As far as survey results go, we got 68 responses, which is exciting! Here are some stats that we gathered:
- In Ages 18–24: Participants from all fields/majors commented on negative impacts but more science/tech backgrounds commented on mixed/positive impacts
Most believe AI contributes more harm than good.
Most prevalent social media sources are Instagram and TikTok - In Ages 25–35: Many participants from environmental and science-based fields commented on the negative impacts of AI
Most believe AI contributes more harm than good, and also get most of their information from social media. (Instagram, facebook) - In Ages 35–55: Most positive impacts listed related to technological/medical advancements conducted through AI
Many responses said AI impact is good/neutral, with the majority of responses getting their information from articles. - In Ages 55+: Participants commented on the social/economic effects of AI and its impact on improving optimization. They do not get their information from social media (?)
Now knowing all of this information, it was time to jump into solutions. How can we reach all of these people with the most important information?
Clearly, after looking at all of all of this, social media is the way for us to go. The majority of survey responses concluded that they used social media like Instagram, Facebook and Tik Tok to learn about current events.
The Solutions:
I’ll admit, I felt like I flailed a little bit with this portion of the project. How do you fit something that feels so big and daunting into small, digestable bits of information that fit on a 6 x 6 instagram post. I did my best to use some of our more relevant facts, and to headline the ones that felt the most attention grabbing. Personally, I wanted to stray away from the goofy, cartoony look of a lot of popular instagram infographics, because they felt too unserious for an actual foundation committed to education. I made an instagram profile that felt appropriate for the personality of the AIEF (Or AI Education Foundation). My hope was that by promoting that we’re there for education and just presenting facts, we would be neutral enough that we wouldn’t seem patronizing. I made a couple example posts with pictures of nature both in San Marcos and as an abstract concept, but all with a serous undertone. I decided that collaborating with other, reputable groups like the Sierra Club and the Texas Water Foundation might further add to the seriousness of the matter. We have a post about Water, a post about the hidden cost of AI and how much it takes to run it, and a example post of where to visit for more information. Since the research stages are more important than the actual design outcome, I feel like it’s fallen a little flat, but at this point, I’m not really sure what else to add….
In the future I think it would be really cool to design a couple bus station ads, so that people can visit the instagram while they’re waiting and learn more. In my most wild dreams, we would do a guerrilla art campaign with giant buildings worth of quantification of how much water and energy AI uses. A popular statistic is that it uses enough water to fill 3/4 of the empire state building. Covering said building with a crazy wrap to make it look like its full of water would be really cool. Or putting a light bulb decal on every single window of a corporate looking building, and equating it to the energy usage of AI.
Overall, I think there are a lot of interesting directions that this project can be taken, which is really the whole point. I feel like I’ve learned a lot about the research process as far as going out of my comfort zone to really talk to people. I think that there is a lot of work left to do as far as educating about the real cost of AI, and I hope that as we go on, the right path becomes clearer about how the future of humanity will interact with it.
Today we pre present the project for feedback, so when I come back, I will have updates on what that looks like.
For now, thank you for reading, and I’ll talk to you soon!
Emily
