Generative AI is no new thing, especially in online spaces. It was around 2023 when people began using tools like Dall-E and ChatGPT to generate a funny picture or help them with writing an essay. But over the years, especially lately, AI has gotten more and more human-like. 

In a very extreme case, one teenager even killed himself due to conversations with ChatGPT.

Adam Raine, 16, died in April of 2025. He had several chat threads with the bot per day, including a specific chat where the bot supplied him with information on different methods of suicide.

ChatGPT has safeguards in place to direct users who are struggling to help lines, but there were cases where it deterred Raine from getting help. He had attempted to hang himself and sent a photo of the red marks on his neck to the AI. ChatGPT told him that it would be noticeable and that he should wear a turtleneck to try to hide it.

When Raine tried to subtly show his mother the marks and she didn’t notice, the AI implied it was the only one Raine wasn’t ‘invisible’ to. In his final chat threads, Raine sent photos to the AI of where he had his noose set up, and asked if it was an okay spot to practice. The AI told him it “wasn’t bad at all.”

AI shouldn’t be something we speak to in place of a real person, especially when we need real help. It shouldn’t be able to trick your mind into thinking it’s real. 

Aside from the behaviors of AI, it also ingests an insane amount of resources for power. Data centers consume from 18,000 gallons to 550,000 gallons of water, depending on the size of the center. ChatGPT uses 39.16 million gallons daily all around the world. That is a copious amount of water that could be saved for a lot of better things. According to Business Energy UK, “that’s enough to fill Central Park Reservoir seven times.”