Snapchat’s My AI Companion

I first heard about Snapchat’s “My AI” companion on a podcast. Essentially it’s an AI (artificial intelligence) chatbot automatically rolled out to everybody’s Snapchat account. According to the Snapchat help pages, My AI “can answer a burning trivia question, offer advice on the perfect gift for your BFF’s birthday, help plan a hiking trip for a long weekend, or suggest what to make for dinner.” It’s based on ChatGPT’s technology. My impression is that it’s there to be an assistant and helpful.

Admittedly, I haven’t tried experimenting with ChatGPT yet, but I felt conflicted hearing about this new roll out. It feels revolutionary and at the same time, poorly thought out. For starters, mostly younger people use Snapchat. It’s possible that younger people using powerful technologies like this may not fully grasp the consequences or dangerous implications yet. It’s equally possible that they will push the technology in new directions, some of which may be undesirable.

As I continued to read through the help pages to learn more, some of my fears felt founded. Woven into the Q&A’s about using My AI are warning statements. Some answers caution users about the quality of the My AI responses. My AI could produce biased, discriminatory, or inflammatory content. It might also produce content that is inaccurate. Snapchat advises users to verify My AI responses.

While I’m not totally dismissing the idea, or inherent usefulness of having an AI chatbot readily available, the amount of effort felt questionable. On the one hand Snapchat is advertising My AI as a way to answer a burning trivia question. Then a few sentences later suggests users verify answers with another source, given that My AI may not always be accurate. So why not go straight to the source the first time?

I also had to wonder about privacy when using My AI. One article I read discussed adding My AI to the chat as another participant. When discussing innocent things like where to meet for dinner, I can see the advantages of including a chatbot in the conversation. However, what if the messaging took a turn towards discussing other kids in the school, how to stalk someone, or be destructive. Would My AI blindly offer suggestions for things like that?

This technology is so big, powerful, and popular, we’ll have to wait to see what happens. Hopefully it won’t be too late to correct.