«We believe that AI can give us new ways of addressing problems and meaningfully improve people’s lives». So says Brigitte Hoyer Gosselink, Head of Product Impact at Google.org and the creator of the AI Impact Challenge, «an open call to nonprofits, social enterprises, and research institutions around the world to submit their ideas to use AI to help address societal challenges».
Announced by Google.org in October 2018, the AI Impact Challenge will fund the best projects, that will have access to a $ 25 million fund. The call expires on the 22nd January and winners will be announced this spring.
Brigitte Hoyer Gosselink has spoken to Vita about this global scale initiative.
How does the Google AI Impact Challenge work and what are its aims?
We launched the AI Impact Challenge to inspire and empower more social innovators to apply AI in their work. The AI Impact Challenge is an open call to nonprofits, social enterprises, and research institutions around the world to submit their ideas to use AI to help address societal challenges. Applications are due January 22 and will be reviewed by Googlers and data science nonprofit DataKind. Final decisions will be made by our international panel of expert reviewers from within and outside Google, with winning organizations announced this spring at Google I/O.
Do you think the Italian nonprofit sector will come up with innovative ideas in the area of AI for social good?
Absolutely! One of the reasons we launched the AI Impact Challenge is because we know that organizations all around the world will have innovative ideas based on the challenges they see in their local communities. For example, university students in Los Angeles, CA recently developed a model using Google’s TensorFlow to identify potholes, road cracks, and other roadway issues, to help inform officials where repairs are needed. Or we’ve seen work within and outside of Google to use AI to help in crisis -- forecasting floods in Northern India where monsoons claim many lives each year, predicting wildfires using biomass measures, and even better earthquake aftershock prediction. From local issues to national or global issues, I’m sure that nonprofits in Italy will have interesting ideas based on their own experience. Google.org has already been working with Italian nonprofits to provide digital skills and promote internet safety so we’ve seen firsthand the great work that is happening in the sector.
What will winners of the Google AI Impact Challenge receive?
We look to bring the best of Google’s resources to the organizations we support. Winners will receive funding from Google.org from a $25M pool, expertise from Google AI, and credit and consulting from Google Cloud. They’ll also participate in a Launchpad Accelerator program to get hands-on coaching and mentorship from experts within Google and beyond.
How can nonprofits incorporate artificial intelligence into their work?
We’re seeing applications of AI across a variety of domains and sectors. Research that we worked on with McKinsey showed that AI has the potential to contribute to all the UN Sustainable Development Goals. For organizations who are interested in incorporating AI but aren’t sure how to get started, we launched a guide on our site (https://ai.google/education/social-good-guide/) that helps explain AI and the types of problems it is well-suited to address. We don’t believe that AI Is a silver bullet so it’s still important that organizations are starting with the problem and then determining whether AI can be part of the solution.
How can AI give new ways of addressing the important challenges of our time and improve people’s lives?
At Google, we believe that AI can give us new ways of addressing problems and meaningfully improve people’s lives. We’ve already seen AI’s potential to transform our own products. And we’ve been applying our own research and engineering expertise to projects that benefit the world through applications like flood forecasting, monitoring whale species, and earthquake aftershock detection. At Google.org, we want to empower organizations on the front lines of solving today’s biggest challenges to apply AI to the problems that they know best. Since 2005, Google.org has invested in innovative organizations that are using technology to build a better world, and we’re excited about AI’s potential as part of the next chapter in this work.
Last year, Google worked with McKinsey to analyse the possible uses of artificial intelligence for the third sector. This research showed that AI could contribute to all 17 Sustainable Development Goals of the United Nations. Can you comment this?
We worked with McKinsey to map out the potential of AI to have social impact and were excited to see these results about the wide applicability, health and hunger, education, justice, equality and inclusion. This is just mapping the potential so the report also notes that we still have to overcome a number of barriers in access to talent and access to relevant data to help us achieve that potential. The AI Impact Challenge is our first step in providing resources to empower organizations to overcome some of these barriers to using AI in their work.
Can you give some examples of nonprofits already implementing AI projects which are helping them achieve their goals?
Resettlement and integration of refugees and migrants is a challenging issue in many parts of the world right now. The choice of where to help refugees resettle can affect their success in adapting to a new place. International Rescue Committee is working with Stanford University to test a predictive model for an individual’s success in a particular city based on factors like demographics, local market conditions, individual preferences and outcomes. Another example is Thorn, an organization founded to stop child sexual trafficking and abuse. Thorn use AI to analyze the content of online ads to identify clusters of suspicious activity. With AI and resources from Google.org, Thorn has been able to identify to law enforcement close to 9,000 traffickers and 28,000 victims. We also embedded a team of Googlers with the organization for six months to support Thorn’s work. A final example is ZSL's application of AI to monitor biodiversity. Google has supported ZSL’s work to use technology to improve their monitoring through remote cameras placed in wildlife areas to capture activity. The ZSL team had to manually review these images to identify species and draw conclusions but is now able to use AI to identify species and free up staff time to focus on interpreting and acting on that information.
Photo of Brigitte Hoyer Gosselink: © Google
Opening picture: pixabay/Gerd Altmann