Editor’s note: On a trip to visit family in Iran, Sophia Kianni made an alarming observation. Despite facing the disproportionate impacts of climate change, and with temperatures in the Middle East rising twice as fast as the global average, her relatives knew almost nothing about the world’s environmental challenges. When she realized that scientific literature wasn’t readily available in Farsi, she took it on herself to translate crucial documents, helping her relatives understand the threats of climate change and the opportunities for action. This inspired her to establish Climate Cardinals, which makes climate resources more accessible to non-English speakers. In this blog post, Sophia and her colleague Hikaru Hayakawa from Climate Cardinals talk about the importance of accessible climate education and the empowering potential of AI technologies in the fight against climate change.
Education is key to climate action, but there’s a barrier that continues to lock billions of people out of the conversation: language. For too long, the international climate movement has been inaccessible to people who don’t speak English. We’re talking about 80% of the world’s population.
Even though only one in five people speak English, more than three-quarters of scientific publications are only ever released in English. Even reports from the Intergovernmental Panel on Climate Change (IPCC), which contain the most comprehensive assessments of how to reduce the impact of climate change, are only available in the six official UN languages. This still leaves out approximately half of the world’s population.
We saw an opportunity to change this. In the first three months of using Translation Hub, with support from the Google Cloud team, we were able to deliver the same volume of translations we had produced in our first two years of operation. This included producing a synthesis of the IPCC report and translating it into over 25 languages.
Climate change is the biggest issue of our generation, so we need to make sure that people have the resources they need to make sense of the disasters that are destroying their communities. At Climate Cardinals, we’re focusing on translation as a vehicle for disseminating this crucial information to the communities most impacted by climate change.
Climate Cardinals launched during the pandemic, immediately earning more than 1,000 volunteers for our cause on the first day. From the start, it’s been a truly youth-led movement: the average age of our volunteers is 16, and we’ve been banging the drum on social media to recruit more young people and expand our executive board.
The enthusiasm of our volunteers keeps our climate mission moving. In our first years of operation, we managed to translate more than 500,000 words of crucial climate information, like climate glossaries and UN documents, into more than 40 languages. We were also able to form partnerships with UN organizations and professional translation networks, like Respond Crisis Translation or Translators Without Borders, for editing and proofreading, helping us to ensure that our translations are correct and credible.
But there have also been challenges. Because our capacity depends on the busy schedules of our student volunteers, we had to be selective about the translation work we could accept. And as students ourselves, managing the high interest and workflows of thousands of volunteers meant constantly adapting and refining our approach, all while staying on top of our other work. So when the Google Cloud team reached out to us, halfway through our third year of operation, we instantly recognized the potential to elevate our efforts and make a greater impact by joining forces.
Initially, a Google Cloud team offered to lend us capacity as part of their social impact initiative. This was a multilingual group in the EMEA marketing team in Google Cloud, who volunteered their time as part of the GoogleServe program. Soon after conversations started, the Google Cloud team saw potential for automation using their AI tools, and started an internal trial of the newly launched Translation Hub product.
Translation Hub is Google Cloud’s AI-powered self-serve translation platform, and it was a real game changer.
Translation Hub automatically translates documents with AI, and then lets us augment and edit the results with human translators. The Hub can process PDFs with design elements and returns a translated document with the same design. It can also translate text within design elements. Initially, we worked hand in hand with the Google Cloud team to integrate the solution into our processes. When we had new translations, we sent them over to the Google Cloud team, who ran them through the Translation Hub. The finished translations then went back to the volunteers in addition to the Google Cloud team, who would proofread, edit, and polish them.
The change in pace was immediate. In the first three months of using Translation Hub, we translated 500,000 words of vital climate information for non-English speakers, for whom there is a drastic asymmetry in available educational materials. But that was only the beginning.
Today, we’re using Translation Hub all by ourselves, and it has become an essential part of our translation pipeline. As a result, we’re more productive than ever. No longer held back by capacity issues, we’ve been able to take on larger, more ambitious projects. That includes documents as long as 200 pages, which would have been tough to manage in the past.
For example, we translated around 100,000 words of climate information into Spanish for Yale Climate Connections to increase their outreach among Latinx communities in the U.S. We also translated documents into more than 20 languages for various global organizations to increase their outreach among youth. These organizations include the United Nations Sustainable Development Network, the Climate Mental Health Network, the United Nations Association Climate and Youth Council, the Environmental Justice Foundation, and the United Nations Committee on the Rights of the Child.
Our busy volunteers really appreciate the helping hand from AI, and so do our partners. According to Yale Climate Connections Editor-in-Chief Sara Peach, “Our partnership with Climate Cardinals has super-charged our ability to make crucial weather and climate change information accessible to our Spanish-speaking audiences.”
Translation Hub has made it much easier for our team to manage incoming translation requests and allocate tasks to proofreaders. The process is now more streamlined, allowing us to be more efficient and conscious of our volunteers’ time and capacity. They can better juggle their volunteer work with school and other commitments, including other forms of climate activism.
As we continue to break language barriers and spread vital climate change information, it’s essential to remember that translation isn’t an end in itself. Our work gets its meaning from its potential to empower and educate people around the world. And while climate change impacts all of us, it has a disproportionate effect on communities that are often left out of the conversation. That’s why it’s crucial to ensure that climate information becomes a public good, accessible to everyone, regardless of language or background. We believe that AI can help make that happen.
With the power of AI, we can create a world in which language barriers in climate information are a thing of the past. We couldn’t do this without the technical expertise of Google Cloud. And Google Cloud couldn’t do it without the passion, community engagement, and grassroots connections of Climate Cardinals. That’s what makes our joint project truly unique. It’s an international, intergenerational, multilingual collaboration between the public and private sector.
Together, we’re offering tech solutions that can bring climate knowledge to the global stage, fostering understanding and inspiring action for a more sustainable future. We hope it’s a model for other youth organizations who are looking for innovative ways to tackle climate change.
TL;DR A conversation with 4o about the potential demise of companies like Anthropic. As artificial…
Whether a company begins with a proof-of-concept or live deployment, they should start small, test…
Digital tools are not always superior. Here are some WIRED-tested agendas and notebooks to keep…
Machine learning (ML) models are built upon data.
Editor’s note: This is the second post in a series that explores a range of…
David J. Berg*, David Casler^, Romain Cledat*, Qian Huang*, Rui Lin*, Nissan Pow*, Nurcan Sonmez*,…