Generic artificial intelligence systems require a lot of energy, but many AI developers hope that the technology will ultimately benefit climate, leading to, for example, a more efficient power grid.
However, the goals of those who are developing AI systems for the climate and those who work on the forefront of climate advocacy are not necessarily in line with. To compare the two groups, University of Washington researchers interviewed nine people developing AI, ranging from graduate students to startup founders, as well as 10 climate advocates, including grassroots activists and environmental nonprofit employees. They found that while developers and supporters all care about the climate movement, their particular values and perceptions differ significantly, especially on topics like ethics.
The team announced their findings on July 8th at the design of the Interactive Systems Conference held at Funchal, Portugal.
UW News spoke with Amelia LeeDoğan, a UW doctoral student at the Information School, and senior author Lindah Kotut, an assistant professor at the Information School, who spoke about the findings of the study.
There are many concerns about the climate impact of AI training and model implementation, but AI can also be useful. How did you find the opinions of the different groups you interviewed diverge from those perspectives?
Amelia LeeDoğan: Most supporters thought AI could potentially be useful in very limited use cases. This includes connecting nature and community members, such as automating existing tasks, creating personas for natural organizations such as rivers and trees, and creating urban agriculture diagnostic tools. There were also some more traditional climate science applications. However, a small number of people were also concerned about not only the climate impact of AI, but also about the kind of labor concerns. AI is not going to fix any issues caused by policy. With these things, climate change is not necessarily a fundamental issue. That's a greater injustice.
What did you learn from this study that surprised you?
ALD: We got many interesting ideas from our supporters and wrote some of them in fictional form for another paper published in January. That's what we call design fiction. You imagine a technological future and analyze it from an academic perspective. Some of the ideas were really original. For example, what if a river could talk to you in English, communicate through purr sounds of water and music?
Lindah Kotut: When I announced this at the meeting, many people were surprised that many developers didn't know what grassroots activists were doing. But activists know what the developers are doing. I feel that if that's turned over, some of the debates about the meaning of AI in the environment will make much progress.
ALD: Many developers in research or non-profit spaces have far less contact with their advocates compared to business people. I asked the developer a very simple question. Have you spoken to the community you are trying to make an impact? Many people said no. Supporters said that if some of these developers appeared at a protest or meeting, they would love it.
Did people have recommendations for solutions to this lack of communication?
ALD: I suggest talking to people as the first step. Advocates want developers to approach their desires and needs. For example, one of the supporters we spoke to was using non-AI tools developed by the government. The government stopped maintaining the tool, which hindered her workflow. That's why she was interested in creating and maintaining new software tools.
LK: There are constraints on both sides. Some developers are working on environmental justice issues as a side project, which limits the time they can spend on it. Advocates are primarily accountable to people and the environment, and are constrained by policy. Therefore, they are thinking about ways to support it through policy changes that impact technology. On the other hand, developers can work in spaces where the profits of the company are against the interests of the environment. There is no solution to that dispute.
When it comes to AI technology, what were the advocates optimistic?
ALD: Many advocates' work is data-intensive and can benefit from automation. For example, look at government databases of PDFs that are not scanned at high resolution. Supporters are also excited about advances in science.
What would you like the public to know about this study?
ALD: Climate change is now. We cannot necessarily wait for AI's promises. We already know that the most effective solution to the climate crisis is fossil fuel logging, policy solutions that protect land and water.
We have found that technological developments that plague social issues and are occurring in the development of climate technology. Many power issues. Developers don't always have the freedom to shape the vision of a big picture for a project. It also extends to developers who may be working on projects for social benefits.
LK: Anyone who can amplify what the grassroots climate community is doing. It brings the change they defend there with the voice they want. Also, in technology, we are very fast. Moved fast, things were this belief. But one of the best foils to not move fast and break things is to listen to support the people you work for. Take advantage of local climate groups and listen to them. Supporting them is an extra step, but listening is the best way to do it. These organizations understand the concern that the community will be directly affected.
Hong Jin Lin, a doctoral student at Harvard University, is a co-author of the study. This study was funded in part by the National Science Foundation and the Department of Graduate Student Equity & Excellence at the University of Washington.
/Public release. This material of the Organization of Origin/Author is a point-in-time nature and may be edited for clarity, style and length. Mirage.news does not take any institutional position or aspect, and all views, positions and conclusions expressed here are the views of the authors alone.
