OpenAI’s big announcement on Monday might be a voice assistant that’s reportedly in development.
According to The Information, OpenAI is working on a technology that rolls audio, text, and image recognition into one product that would say, help tutor kids with their math homework, or “give people information about their surroundings,” like translating or helping with car trouble.
On Friday, OpenAI shared that it would be holding a livestream this coming Monday (notably, a day before Google I/O, Google’s developer conference). While initially expected to be an announcement about a ChatGPT search engine, or the unveiling of GPT-5, CEO Sam Altman shot down those rumors and posted on X, “we’ve been hard at work on some new stuff we think people will love!” That new stuff might be a voice assistant.
Mashable reached out to OpenAI for comment, and will update this story if we hear back.
Last September, OpenAI rolled out voice and image capabilities, claiming “ChatGPT can now see, hear, and speak.” (Sidenote: ChatGPT can’t technically do any of those things, because it’s not a living being, but it can process images and audio in real-time, mimicking human senses.)
In the announcement, OpenAI demo-ed ChatGPT’s ability to troubleshoot problems with a bike seat, and with the help of synthetic voices, casually chat with a user instead of providing text responses. The outlet reported that the new model would bring these modalities together.
The Information noted that a voice assistant is currently too big to run on personal devices, but a cloud-based service could be deployed for customer service agents and could recognize when a customer is being sarcastic.
However, the moment this technology gets into the hands of users is when its transformative potential will become clear.
Speaking of which, OpenAI is reportedly in talks with Apple about integrating ChatGPT with iOS 18, which the public expects to hear more about at Apple’s developer conference WWDC in June. Concurrently, Apple is planning a major overhaul of its voice assistant Siri per The New York Times.
If that weren’t enough, Google is expected to announce Gemini updates at its developer conference, Google I/O on Tuesday. Google is also reportedly in talks with Apple about bringing Gemini to the iPhone. So despite OpenAI’s close ties to Microsoft, reports of a deepening association with Apple — even though Apple is also reportedly talking to Google — hint at the possibility that generative AI is helping to build intertwined super-entities with uneasy alliances.
But whatever happens at Google I/O and Apple WWDC, it comes after OpenAI’s mysterious announcement on Monday.
