Data and AI are revolutionizing the future.
How often have we heard this sentence in the last few years? And how often has it been followed by an explanation of how data and AI are actually being used by the people claiming its revolutionizing powers?
For Dr. Ayesha Khanna, it’s not as simple as making a statement. For her, the first step is to build up a team of like-minded individuals. Working with machine engineers, data architects, and data scientists, many of whom come with a similar academic background, she helps businesses create solutions for problems using artificial intelligence and data.
The second step is the question: what are you going to do with it? For her 021Disrupt talk, she spoke about how she and her team use AI For Good in 3 different industries: healthcare, financial services and governance, or smart cities, and how these industries use artificial intelligence.
For hospitals, how do you optimize their network by being able to predict where the next outbreak of a virus or disease might be? In order to make these predictions, the health ministry needs to be connected to many different data points that can be analyzed. ADDO AI is connecting hospitals and clinics into one database so that it’s easier for the city to analyze the demographics of the patients, their symptoms, how long they contracted Covid-19 for, etc. so that the information can be added to a heatmap of patients. This helps to predict how likely it is for there to be more patients within the surrounding area of previous patients.
In the beginning, AI confirms human intuition. As time passes, it begins to pick up on patterns that we miss out on. For example, if a group of people go to a pharmacy for the same kind of medication every month, or cigarettes, what underlying illnesses can be predicted based on their purchasing behaviors? When you apply machine learning to this kind of data, it’s easier for patterns to emerge.
Here’s the lifecycle of AI: Gather data > Organize data in a repository, known as Data Link > Once the data is organized, the AI can begin its task of finding patterns within the data that humans often overlook. This information allows us to optimize our response time.
When AI reaches the end of its lifecycle and gives us all the patterns, the next step is to try to figure out how to improve your hospital’s response to the data. If you know a certain area is going into the red zone, you can use AI to predict things like how long the hospital’s emergency queue lines will be, and how to handle it. You’ll know which days or weeks to schedule more cardiologists, ER surgeons, nurses, pediatric specialists, etc.
68% of people in emerging markets do not have bank accounts. Can you use alternative data from your digital footprint to assess credit risk?
Take the example of a woman in Karachi who wants to open a beauty salon. She has no education, no previous job experience, no background in business: who’s going to give her a loan?
But if she has an e-wallet, you can monitor her transactions to see whether she pays her bills on time or if she’s scrolling through social media all day, or she’s on YouTube, learning skills related to her job. Does she type fast, or with one finger? Does she type in all caps, lowercase letters or a mixture? Dr. Ayesha relates that as small and irrelevant as they may seem, all these things have some correlation with credit risk assessment.
As a country, we are severely disadvantaged when it comes to financial services. Unless you have a bank account, unless you have contacts in the industry or something along those lines, you’re often not taken seriously. But by using automation and AI, what Dr. Ayesha and her team are doing is taking the bureaucracy out of the equation for people who it’s stacked against.
Technology is always a double-edged sword. The same AI algorithms that can be used to predict your likelihood of developing cancer in the next few years of your life can also be used to generate fake data and fake news that can lead to misinformation.
In Singapore, where Dr. Ayesha is currently based, there exists an AI and Ethics Committee whose sole purpose is to seed out bias manipulation and put policies in place that prevent the invasion of data privacy.
Data and AI should be used to ensure that they’re helping to make someone’s life easier, which is something that Dr. Ayesha and her team are working towards every day.
For the full talk, go here.