When Google opened the call for proposals on how to use artificial intelligence to address problems facing society, it was answered with 2,602 from 119 countries. This tremendous response allowed 25 proposals to be funded with $25 million in grants, support from Google AI experts, and other tools to help turn their visions into reality. Through this process, Google was able to garner important insights into the current state of AI and how to accelerate its use as a tool to solve major problems facing society. The insights gained, and shared, by Google point to important opportunities and areas in need of improvement in order to advance AI for positive societal impact. 

It is essential that organizations learn when to use machine learning. In some circumstances, machine learning is actually more expensive and time-consuming than other methods that would achieve similar results. For the organizations that submitted proposals, some drastically underestimated the complexity of using machine learning to tackle the problem they chose to address, while for others, machine learning has not yet advanced enough to make their solution possible. 

More challenges to face, Google found that organizations tackling issues such as crisis response, economic empowerment, and equality were lacking meaningful data sets. This creates an inability to accurately train AI to classify specific information. Additionally, vulnerable populations of people are more difficult to collect reliable data on, as they are typically highly sensitive to privacy and more often are transient. 

Other sectors, such as health, environment, and education already have access to the data they need, the main challenge being organizations’ ability to access the data. For organizations that already own the data, this presents an incredible opportunity to participate in data-sharing partnerships. 

As AI and machine learning improves, so have opportunities for analysts capable of interpreting and engineers to run through the data sets. Machine learning libraries and open-source tools have made it easier for organizations to utilize AI, resulting in a decreased need for dedicated AI specialists, and an increased demand for data analysts and engineers. 

Organizations could greatly benefit from partnerships in order to effectively implement AI for positive social impact. For example, organizations that have extensive expertise in their sector can often be aided by organizations that have the technical knowledge and background to implement AI. During their review, Google found that the organizations with established partnerships had developed the more comprehensive applications and were in the best position to utilize AI for good. 

Additionally, Google identified many projects that were similar, with overlapping problem areas. The organizations focusing on shared problems could likely benefit greatly from shared resources. If organizations worked together on the common issues, it could easily accelerate programs and make positive changes in society. 

By highlighting challenges and opportunities faced by organizations working to implement AI for societal growth, Google provides valuable information and insights. Hopefully this information allows organizations to better collaborate and advance AI, and provide long-lasting solutions to current problems.