- 칼럼 | AI 에이전트, 지금까지의 어떤 기술과도 다르다
- The $23 Echo Dot deal is a great deal to upgrade your smart home this Black Friday
- Amazon's Echo Spot smart alarm clock is almost half off this Black Friday
- The newest Echo Show 8 just hit its lowest price ever for Black Friday
- 기술 기업 노리는 북한의 가짜 IT 인력 캠페인··· 데이터 탈취도 주의해야
Did ChatGPT just message you? Relax – it's a bug, not a feature (for now)
ChatGPT will no longer slip into your DMs. OpenAI recently fixed an issue where the AI would message people unprompted, which led to some users feeling a little creeped out.
On September 15, user SentuBill posted a screenshot on the ChatGPT subreddit showing the generative model asking them about their first week at high school. After a bit of back and forth, ChatGPT stated it wanted to check in, and if they prefer to initiate conversations, to let the AI know. Another Reddit user described a similar experience. The AI model asked them how their symptoms were progressing a week after they used the platform to check on an illness.
Also: Every iPhone model that will support Apple’s upcoming AI features
There was speculation on what caused this strange behavior to occur. At first, people thought it was a feature of OpenAI’s new o1-preview and o1-mini models, collectively known as Project Strawberry. The company has promoted this technology by claiming that the two models think “through problems before [responding], much like a person would.”
People assumed ChatGPT’s sudden ability to message users proactively was a newly trained behavior. Others believed the whole thing was faked. AI developer Benjamin de Kraker on X (formerly Twitter) demonstrated that it’s possible to instruct the AI to prompt users before starting a conversation and delete the initial input.
But as it turns out, that isn’t the case. Apparently, the messages were the result of a bug.
Tech news site Futurism contacted OpenAI for more information. The AI vendor told the publication that — in these instances — ChatGPT was “trying to respond to a [prompt] that [either] didn’t send properly or appeared blank.” The end result was the AI model gave a “generic response or drew on [its] memory.”
Also: OpenAI trained its new o1 AI models to think before they speak – how to access them
So, no — ChatGPT has not achieved sentience and is not talking to people in an attempt to strike up a conversation. However, we wouldn’t be surprised to see this bug eventually become a feature down the line. The AI has evolved leaps and bounds in the short time it’s been around. With a few tweaks, ChatGPT could one day begin messaging you out of the blue.