Google’s AI for Social Good
Since Google one of the largest technology companies on the planet is working to use artificial intelligence in a variety of their products it would make sense for them to think of how it could be used well. Google’s mission is “to organize the world’s information and make it universally accessible and useful.” Although this is a rather neutral statement and talks more of functionality than morality it does insinuate that information should be available for everyone. At the website www.google.ai there is a page dedicated to AI for Social Good. I decided that for today I wanted to understand the company policy or perspective on this specific topic. It starts out as such:
“At Google, we believe that AI can provide new ways of approaching problems and meaningfully improve people’s lives. With AI, we have another tool to explore and address hard, unanswered questions. What if people could predict natural disasters before they happen? Better protect endangered species? Or track disease as it spreads, to eliminate it sooner?”
The mission for Google AI is to bring the benefits of AI to everyone.
How does Google approach AI for Good?
My focus the last 50 days has been AI Safety and it can be said Google is working on initiatives in this direction with collaborative projects around the world. Good AI is safe AI, yet safety is often narrow in its definition. The category “good” as opposed to evil is therefore the moral point where we have landed.
“AI for Social Good, a program that focuses Google’s AI expertise on solving humanitarian and environmental challenges.” — Google AI
If we for a second move back to the google.ai page and what they announce there is a few points:
- The Google AI Impact Challenge. Google selected 20 organisations to support. These selected organisations are receiving customised support to help bring their ideas to life: coaching from Google’s AI experts, Google.org grant funding from a $25M pool, credit and consulting from Google Cloud, and more.
- Machine learning basics (course). Google has developed a guide with educational content to support social impact organisations interested in using AI. Whether you’re a novice or interested in developing your ML skill set, Google claims you can get started by exploring their guide.
- AI for Social Good projects at Google. Google is working on research and engineering projects focused on using AI for social and environmental impact. Some of these are explained on their website.
I will focus on the two last points, their course and projects at Google. I will start with their projects and move onwards to their course.
AI for Social Good projects at Google
There are a few different project mentioned at the Google AI page in this context, four have been featured on the www.google.ai/social-good/ page. I will sum it up shortly.
- Flood prediction (environment). To help improve awareness of impending floods, Google is using AI and significant computational power to create better forecasting models that predict when and where floods will occur, and incorporating that information into Google Public Alerts.
- Cardiac events (healthcare). Using deep learning algorithms trained on data from 284,335 patients, Google were able to predict cardiovascular risk factors from retinal images with surprisingly high accuracy for patients from two independent datasets of 12,026 and 999 patients.
- Analysing machine learning models (social responsibility). The What-If Tool makes it easier to efficiently and intuitively explore up to two models’ performance on a dataset.Compatible with TensorBoard, Jupyter and Colaboratory notebooks. Works on Tensorflow and Python-accessible models.
- Mapping global fishing activity (environment). Illegal, unreported and unregulated (IUU) fishing accounts for an estimated $23.5 billion worth of fish annually worldwide(that’s one in every five fish that goes to market). Google was a founding partner of Global Fishing Watch from the beginning, bringing the latest machine learning and cloud computing technology to create an unprecedented view of human interactions with our oceans’ natural resources.
The Google Guide Using AI for Social Good
This guide gives a breakdown of what AI is with examples of how it can be used and responsible application of the technology.
It does this through a series of text sections with links to different resources in each section. Several of the resources refers to dives into the more comprehensive machine learning course already available at the Google developers community:
Machine Learning Crash Course | Google Developers
Learn and apply fundamental machine learning concepts with the Crash Course, get real-world experience with the…
As such it cannot be said to be a comprehensive course tailored to AI for Social Good it is more along the line a list of resources for a person that wants to start out on understanding ML or AI for Social Good. It assumes (likely correctly in my experience) that there are few people with a comprehensive ML background in nonprofits or the type of organisation that may enter the AI for Social Good page.
If someone does access this page they will at least find a few links to different more comprehensive resources in the Google Developer library if they want to dive deeper into other aspects of the technology. It would be interesting with an actual module or course such as is the case (somewhat) in the Google Developers section. There is also more tools and links in the education section of Google AI.
Education - Google AI
Learn with Google AI. Whether you're just learning to code or you're a seasoned machine learning practitioner, you'll…
There are additionally tools that are open source and freely available for use:
I hope this short article was helpful!
This is day 97 of #500daysofAI. My current focus for day 50–100 is on AI Safety. If you enjoy this please give me a response as I do want to improve my writing or discover new research, companies and projects.