- OpenAI filed a trademark application for a "voice engine" and to build "digital voice assistants."
- The application came a day after Sam Altman hinted at other releases before the upcoming GPT-5.
Sam Altman might be striving to take on Apple's Siri and Amazon's Alex voice assistants.
OpenAI has filed a trademark application to build "digital voice assistants" and a "voice engine", which signals that it may be set to release a new product.
The application was filed with the US Patent and Trademark Office on Tuesday, a day after an interview with Sam Altman was broadcast in which he said OpenAI has "a lot of other important things to release" before its upcoming GPT-5.
Those features might not come to fruition as companies often file trademark applications for ideas that never see the light of day. OpenAI is, however, expected to release a "materially better" upgrade of its ChatGPT model mid-year, as Business Insider previously reported.
In an interview with podcaster Lex Fridman, Altman said: "We will release an amazing new model this year. I don't know what we'll call it. We'll release over in the coming months many different things, I think they'll be very cool."
He added: "I think before we talk about a GPT-5 like model called that or not called that or a little bit worse or little bit better than what you'd expect from a GPT-5, I think we have a lot of other important things to release first."
The company doesn't yet offer a digital voice assistant to its users, although it does have an API that converts speech to text called TTS, according to its website. It also has a general-purpose speech recognition model called Whisper.
OpenAI also filed trademark applications for its future models, including GPT-6 and GPT-7 in October.
The filing for GPT-6 includes simulating conversations, sharing datasets for the purpose of machine learning, predictive analytics, and "analyzing algorithms that are able to learn to analyze, classify, and take actions in response to exposure to data."
The GPT-7 trademark application covers software using artificial intelligence for music generation, conversion of text and data files into software code, and creating and generating software code. Both applications are still under examination.
The Patent Office might not grant those trademarks, though, as in February, it rejected OpenAI's attempts to trademark "GPT." The agency said it's a "widely used acronym" that's "merely descriptive" as it means "generative pre-trained transformers."
The "voice engineer" trademark application, which is pending a review from an examiner, intends to cover software for the following 10 areas:
Building digital voice assistants
Voice and speech recognition, processing voice commands, and converting between text and speech
Processing voice commands, and converting between text and speech
automatic speech and voice recognition and generation
Generating voice and audio outputs based on natural language prompts, text, speech, visual prompts, images, and/or video
Generation of audio and/or voice in response to user prompts
Machine learning-based natural language and speech processing
Multilingual speech recognition, translation, and transcription
Using artificial intelligence for automatic text-to-voice and text-to-audio conversion
Development of voice service delivery
OpenAI didn't immediately respond to a request for comment from Business Insider.