Some Artificial Intelligence technologies have been around for more than 50 years, but advances in computing power and new algorithms have led to major AI breakthroughs in recent decades. AI generates a lot of hype and excitement since it’s still associated with a vision of the future. Meanwhile, we already rely on it every day. Do you use Siri or Alexa? Do you use spam filters on your e-mail? Is there an algorithm that shows you recommendations from Netflix or Spotify? How many times have you had a conversation with a bot that tried to solve your problem with a device or service? AI is all around us. It’s used commercially by companies worldwide for years, but now a new trend has emerged. AI for good is a global trend — many different companies, like Microsoft, Google or Amazon, are trying to solve the world’s most pressing challenges with the use of technology. Other organisations, like the United Nations, look to AI for ways to accelerate their sustainable development goals.
What is AI?
An AI system combines and utilises machine learning and other types of data analytics methods to achieve capabilities traditionally associated with intelligence. This refers not only to computer science but also psychology, linguistics and other fields. Many different definitions highlight the ability to plan, reason, learn, sense and communicate in natural language.
Why is it beneficial?
AI can help us get insights from large amounts of data — understanding it and learning from it. It sees patterns where we don’t and connections that would be impossible for the human brain to make.
There are two main ways we deal with large volumes of data to get some insights. One is to organise them through smart visualisation, which provides an overview otherwise obscured by the vastness of the data. The second is to apply machine learning algorithms, which analyse vast amounts of data faster and more accurately than the human brain, providing inferences and sometimes even suggesting solutions.
When to apply it to the civil sector?
You have a database that is too big to be processed manually
Imagine you are trying to collect data about COVID-19 patients to establish the usual length of the incubation period. In the beginning, with just a few cases — you can try and do it yourself. Once you fill a whole spreadsheet — you might need more advanced help. And once the spreadsheet is no longer enough, you’re likely to require some very specialized assistance and professional tools even to store the data, let alone perform an analysis of it. This problem is even more profound when you deal with images or video data, as you can quickly exceed the capabilities of your local computers.
Your data comes from many different sources
Let’s say you receive statistics describing (among many other things) the length of the incubation period of the SARS-CoV-2 virus from WHO, NHS, local hospital and directly from some patients. Data analysis can help you sift through all those quickly and efficiently and retrieve the exact information you are looking for. If you receive this data regularly, the whole process can be automated, streamlining the information flow and significantly shortening the time between data collection and analysis.
You are trying to find answers to some complex questions
If you would like to know whether people with diabetes are more likely to suffer from COVD-19, you need to consider many different factors, like age, sex, ethnicity, weight, coexisting conditions, and many others, before making any conclusions. Again, this is a job for algorithms that are suited to take into account many different variables.
Your database comprises different types of images
Using tailored machine learning algorithms will help you make much faster and more detailed inferences than a human brain could when classifying images. This is the approach taken by Appsilon when working on Mbaza AI — a machine learning model which classifies images from camera traps taken in national parks in Central Africa. Spotting and classifying animals visible on those cameras used to take months or sometimes years. Now, it takes days and can be done offline on a standard laptop.
These are just some basic examples of use cases. If you want to explore more opportunities for NGOs connected with AI, please join the free webinar “AI for Good” on the 28th of April 2021, organised by Tech To The Rescue, a non-profit that matches tech companies with NGO’s to solve the world’s problems.
More information: https://www.eventbrite.co.uk/e/ai-for-good-registration-151034112095 .
