Benefits of AI

- There are many beneficial uses of AI. Some intended to help specific groups of people and others intended for the general population. In this video, I'll share a couple of examples of how AI can benefit people in their everyday lives. At age 30, Akiko Ashii lost her eyesight following a surgery. Fearing that she would lose her independence, she sought out technology options that would allow her to more easily navigate the world around her. One of these technologies was Microsoft's Seeing AI. Seeing AI is an intelligent camera app that utilizes computer vision to translate the visual world into audio. It has helped people who are blind or who have poor vision with more than 20 million tasks, such as recognizing products, people and currency. By using Seeing AI, Akiko can now more easily navigate new spaces, read handwritten notes and printed text and even recognize her daughter Ami's facial expressions. To her, seeing AI provides a kind of independence that is priceless. From this example, we can see the tremendous capacity of AI to assist individuals. Let's consider now in more detail how AI can generally can enable greater efficiency, effectiveness and equity. AI is well-equipped to perform repetitive and large scale tasks at a much faster pace than humans. Take, for example, credit card fraud. Between 2019 and 2022, the Federal Trade Commission saw reports of credit card fraud increase by 38%. This is due in large part to credit card companies implementation of AI that can track large numbers of credit card fraud risk indicators, such as the frequency and location of transactions. Historically, this type of fraud has been corrected retroactively, only after a fraudulent transaction has been processed. Today, AI is increasingly used to efficiently evaluate transactions and make billions of decisions in real time, stopping a fraudulent transaction before it is processed. While efficiency is beneficial, it's also important that AI be effective. Continuing with the example of automated credit card fraud detection, legitimate transactions falsely flagged as fraudulent are harmful to credit card companies and consumers. With better data and models, companies can more effectively identify and prevent fraudulent transactions and protect customers from unwarranted account freezes. In addition to efficiency gains, AI can also support greater equity. Let's consider credit reporting and scoring systems that are used to determine eligibility for a credit card. Traditional credit reporting and scoring systems have been found to perpetuate bias. New AI driven models of determining credit worthiness are now being developed to identify and address hidden proxies for discrimination, such as correlations between debt load and housing status with protected characteristics like race, gender or age. Advancements in the use of AI models can help to uncover such hidden biases and support more equitable access to credit. While artificial intelligence can enable greater efficiency, effectiveness and equity, it can also undermine these goals. As we transition to the next video, I encourage you to think about ways that AI can benefit you, your profession or your community and the potential risks it may pose.