SAFETY TIP OF THE WEEK

“A computer can never be held accountable, therefore a computer must never make a management decision.” Know when you are using apps that utilize AI and make informed decisions about what information you share with those apps.

TODAY’S TOPIC

Artificial Intelligence: Illuminating the Black Box

When you hear words like artificial intelligence (AI), algorithm, or ChatGPT, maybe you wonder how exactly these tools work. Maybe you feel slightly overwhelmed because you hear about AI all the time, and you see its products, but you couldn’t confidently say how it works. You wouldn’t be alone! Now that AI applications are used in lots of different digital processes, it's important to understand how those tools work and what their dangers are. 

One of the biggest problems with AI tools is that they don’t always show where their data comes from, or how they made their conclusions or products. AI models like this are often called “black box” models, as they hide their processes from users. A black box AI application’s internal workings can be a mystery not only to users, but to the very people building the AI application as well. Whether these models are intentionally designed to protect intellectual property or become black boxes as a part of their training, they can leave users and developers feeling wary about their applications. 

But what is the real danger of these AI applications? Depending on how you use these applications, the danger varies. For most applications, AI can collect your data and companies can sell that data in order to make targeted advertisements on your other applications. For social media, AI can target users with content that drives up user interface, making apps like TikTok and Instagram more enticing and harder to put down. For example, Instagram and TikTok have both used algorithms to elicit intense feelings of anger, as messages that make us angry are more likely to get us to react and engage with these apps. With more data, AI image generators can create deepfakes that look shockingly real. These images and videos can become tools of misinformation targeting you and your child. 

_{WHAT CAN YOU DO?}

So, what can you do to avoid giving out personal information, and how can you identify deepfakes? Here are some tips for protecting yourself and your child from the dangers of AI applications:

• Know When You Are Using Apps that Rely on AI: AI applications are used so often in apps in ways that you may not realize. These applications can collect your information to be sold or to target content toward you and your child. 

• Practice Identifying DeepFakes: As deepfakes become more and more difficult to spot, some organizations have created quizzes and practice games to help users recognize deepfakes. Take some of the quizzes with your child to help them spot deepfakes too!

• Verify Information if It Seems Suspicious: If you see a potential deepfake, check other sources to be sure the information it is presenting is accurate. 

The IBM training manual from 1979 states, “A computer can never be held accountable, therefore a computer must never make a management decision.” AI applications are complex tools, but they are still imperfect tools. We all need to be responsible for the decisions we make, and in this modern world it is essential for us to know the imperfections of AI tools and when they are being used. If not, these tools may come to shape our entire reality. 

▶ Slang Word of the Week: “IYKYK” – literally “if you know, you know,” often used when posting something to a large audience for a select audience. It can be a shady way of talking to everyone behind someone’s back, or instigating a fight with someone by being shady.