OpenAI says ChatGPT messages were initially a bug, not a new feature
ChatGPT’s latest o1 model – or Strawberry as it’s codenamed – has started rolling out, promising to make its chatbot AI more human than ever. That includes the ability to better understand natural language, but few expected ChatGPT to be the first to be able to send you a message. For a brief time, it seemed it could, although that was apparently a bug rather than a hint at a new upgrade.
Several users reported on social media that ChatGPT had started a conversation with them; in an example of this is from Reddit ChatGPT asked the user how their first week of high school had gone; when asked if they messaged first, the Open AI bot proudly replied “yes,” later expanding that it can now initiate a conversation to follow up on issues it had previously discussed. Another user in the comments claimed that the AI asked them about health issues they had previously discussed with it.
As proof of the new feature, the user can also shared a link to their conversationto prove that ChatGPT did indeed initiate the conversation without asking. While some were skeptical, with one Twitter user which shows how the chat could be fake.
Either way, people were excited about these types of features, real or not. After your first day of high school or after your health check, there’s a (slightly creepy) sense of human compassion, but we think of the more mundane memories.
How many times have you done research for an upcoming trip, for example, only to forget everything you researched? ChatGPT can now remind you to book a table at that restaurant it recommends or plan your visit to that must-see famous landmark, so your planning won’t be in vain.
However, it appears this was a bug and not an intended addition to ChatGPT, which is currently being A/B tested.
Gone but not forgotten
OpenAI has now responded, saying: Futurism in a statement that it had “addressed an issue where ChatGPT appeared to be starting new conversations”, adding that “this issue occurred when the model attempted to respond to a message that was not sent properly and appeared empty. As a result, it would either provide a generic response or call on ChatGPT’s memory.”
That means ChatGPT can’t start a conversation, at least for now. But given the buzz this situation has created, we wouldn’t be surprised to see OpenAI start working on something similar.
If the bot already has the ability to recall events from previous conversations, there’s no reason why a little extra logic programming wouldn’t allow it to start chats based on what it knows – provided you give it permission. For now, we’ll just have to wait and see where OpenAI takes ChatGPT.