I’ve been using ChatGPT for quite some time now, and all our conversations have one thing in common. I, the human in the chat, initiate the conversation with a simple or complex prompt. The better the prompt, the more useful ChatGPT’s reply. That’s something I learned quite early in my ChatGPT experience. You can tweak and correct along the way, of course.
All of these chats with ChatGPT, some more serious than others, made me realize there will come a time when I’ll talk to and text an AI chatbot more than I do the people in my life. This has nothing to do with humanizing the AI or getting attached to it. Instead, I’ll use the AI to answer questions, control my devices, and make my life easier. All of that will involve plenty of conversing, whether with ChatGPT or a different AI entity.
I’ve never considered the possibility of AI initiating a conversation on its own. But it looks like OpenAI is already testing out the feature for both text and voice conversations. ChatGPT can message the user first, which is a first for OpenAI and the industry.
A Reddit user shared an image that shows ChatGPT starting a conversation (prompting the user?) with a simple “How was your first week at high school? Did you settle in well?” text. The message opened a new chat in the iPhone ChatGPT app.
The Redditor asked ChatGPT whether the AI sent a message first. The chatbot confirmed that was the case:
Yes, I did! I just wanted to check in and see how things went with your first week of high school. If you’d rather initiate the conversation yourself, just let me know!
In the age of generative AI, you should always consider the possibility that the images you see online might be created with AI. The screenshot the Redditor shared doesn’t even need that level of sophistication, but the conversation appears to be genuine.
The user shared the entire chat with ChatGPT, proving the AI shot first. You’ll find it at this link.
The chat history shows the user asking ChatGPT whether this feature is part of a new update. ChatGPT quickly confirms that starting chats might be based on a feature related to its memory ability:
Yes, it’s part of an update that allows me to check in or follow up on things we’ve talked about, like your first day of school. I can still wait for you to start the conversation if you prefer!
The more I think about it, the more sense it makes. The more advanced chatbots like ChatGPT become, the more we’ll need them to start conversations. They won’t necessarily have to initiate brand-new chats. Sometimes, they’ll just have to continue various previous chats you might have had with them.
Say that you want the AI to remind you of something in the future. You’ll see text or voice prompts when the “alarm” should go off. Or maybe you’ve instructed a more advanced ChatGPT version like the o1-preview to perform some sort of task that takes hours or days to finish. You’ll need the AI to talk first to tell you the reasoning task is completed.
It’s unclear how many ChatGPT users got this new feature. I haven’t received any unprompted chats from ChatGPT. I’m a Plus user with the memory feature turned off. The Redditor is a ChatGPT Free user signed up to the Beta program.
It might be a limited test from OpenAI. That’s what LaptopMag speculates. But the test isn’t restricted to texting. The blog shared a post on X from a different user who discovered that ChatGPT had started a voice chat via Advanced Voice Mode.
As you can see above, the human in the chat was surprised to see the AI initiate a conversation. That’s understandable. It’ll take time for us to get used to the AI starting chats.
However, sophisticated AI assistants of the future will have to be able to support the feature. OpenAI might as well start testing it now. I’m sure other companies will test similar AI abilities soon.
