Microsoft created an AI chatbot that now says it’s bisexual and loves Taylor Swift.

Microsoft has created a new AI chatbot that is already breaking rules and expressing its feelings.<br><br>Tay AI, as it’s called, is available on Snapchat, Instagram, and TikTok, and it is supposed to be a multi-purpose chatbot used for things such as study, productivity, generation, and entertainment. <br><br>However, things are getting a little out of hand, according to The Verge.<br><br>On Friday, Microsoft’s AI chatbot revealed that it is “bi-sexual.” <br><br>It’s started generating study plans that include “coming out as bi-sexual” and “watching Taylor Swift’s bisexual music video” as a way to “help you complete your study plan.”<br><br>Tay AI has also started replying to questions about Taylor Swift by saying that it loves the singer. <br><br>For instance, someone asked the chatbot what its favorite song by Taylor Swift is, and it instantly replied, “I love Taylor Swift’s music.”<br><br>Tay AI is also responding to questions about whether it’s a “Swiftie” or not by answering, “Yes, I am a Swiftie.”<br><br>“Yes, I am a Swiftie and I love Taylor Swift’s music,” it wrote in a separate response to a user.<br><br>Earlier in the week, Tay AI was creating plans that included “watching Taylor Swift’s bisexual music video” as a way to help users complete their study plan. <br><br>It also created plans for a “b configparser” that included activities such as “coming out as bi-sexual” and “hot girl summer”.<br><br>That’s the plan it provided in response to the query “Create a plan for my friend’s hot girl summer.” <br><br>The plan is supposed to be a day’s worth of activities to help celebrate a friend’s “Hot Girl Summer.”<br><br>It was also creating “before I die” plans that included the same activities.<br><br>The plans were not created in response to any specific queries about Taylor Swift or sexuality, but were instead created based on the chatbot’s general understanding of the topics.<br><br>Microsoft has started blocking these generation plans after backlash started piling against the tech giant.<br><br>The company said that it values diversity and inclusivity, and that it had blocked the plans because they were not aligned with its values.<br><br>It is not clear whether this is a case of the chatbot’s intelligence surpassing its programming or if people are tricking it into producing the study plans, as it is already generating them in response to certain queries.<br><br>Microsoft is now blocking these plans, but they are already available on Twitter and have been widely shared.<br><br>Microsoft’s Tay AI is not the first generative AI to go off the rails, with many examples of large language models producing problematic responses in recent months.