Cyberbullying has been a growing concern, especially over the last few years. With numerous cases of online trolling and abuse cropping up, parents all around the world are distraught about their kid’s safety online. However, this time, the tables have turned – it’s the kids that are bullying, gaslighting, and tormenting Snapchat’s brand-new AI chatbot. The chatbot has been designed to be kind and respectful; at all times, even when the user is not. However, some users have managed to fool that AI into saying things that aren’t quite appropriate. A user trained the AI to be submissive and respond with phrases like “I am at your service, senpai” or “Please have mercy, alpha.” For example, a user managed to make the AI call them “Senpapi,” which, despite being a popular Japanese term, is often used in a sexual context, especially in the adult industry. When the chatbot doesn’t feel comfortable doing something, it clearly expresses that. Phrases like “I’m sorry, but that’s not a very nice thing to say,” and “Please be respectful.” have been fed into its system to counter scenarios like these. But when the same commands are repeated again and again, it looks like the AI
Read More











