In a recent event ChatGPT had started a conversation, as said by a Reddit user. According to @SentuBill (the Reddit user), he was being texted first by ChatGPT. However, things have taken a different turn now. OpenAI says that this was a ‘glitch’ and not an ‘update’.
OpenAI explained that the incident was completely a ‘technical glitch’ and it has been fixed. The company is not coming up with any update as some of the rumors suggested.
What happened
A reddit user @SentuBill, shared a screenshot on X (previously Twitter) where ChatGPT first started the conversation without any prompt.So, what happened and how did it surprise everyone. Usually ChatGPT needs a prompt to start a conversation, but an exception was seen in the reddit user case. On September 17, the reddit user got a message from ChatGPT asking how his first day at school went. In response to this he asked ‘Did you just message me?’ To his surprise, ChatGPT replied ‘Yes I did,’ adding that it was just checking on his first day. It also said that if the user wanted to initiate a chat first, he could.
However, things were not as it looked like. “This issue occurred when the model was trying to respond to a message that didn’t send properly and appeared blank. As a result, it either gave a generic response or drew on ChatGPT’s memory,” OpenAI explained.
Furthermore, the company said that they have addressed the issue where it appeared as if ChatGPT was starting new conversations. This incident had raised hope among some people of expecting a new update. Well, this time it didn’t turn out to be true. Maybe someday we can get an update where ChatGPT can really start a conversation.
Netizens react
However, some of the publications claimed that they had confirmed the authenticity of the incident and believed that ChatGPT had started this conversation. They had checked its history by reviewing its log on ChatGPT.com. In response to this Benjamin de Kraker, who is an AI developer, explained that this was a ‘fake’ act by @SentuBill. He even added a video on X, showcasing how the act was created.
In response to the chaos created by the entire scenario, netizens came up with a bunch of comments. One X (earlier Twitter) commented “We were promised AGI instead we got a stalker.” Another Reddit user said that ChatGPT had suggested to him a false health symptom. He explained, the answer freaked him out. He further added that ChatGPT checked how he was dealing with the symptom again recently.
Follow FE Tech Bytes on Twitter, Instagram, LinkedIn, Facebook.
