OpenAI might have Apple, Amazon and Google in its sights for its subsequent huge push in synthetic intelligence, taking up the voice assistant market with a brand new Voice Engine device.
Whereas ChatGPT has a voice-friendly interface on cell units (and just lately launched a means so that you can converse your responses on desktop), new trademark application by OpenAI for the phrases Voice Engine relates particularly to the creation of digital voice assistants.
It’s now potential to vary the default voice assistant on Android. Apple seems to be in talks with various AI firms about the way forward for AI on the iPhone, so this could possibly be a preemptive transfer by OpenAI to faucet into a possible new market.
Apple can be rumored to open a devoted AI app retailer with the subsequent main replace to iOS, which might create a brand new marketplace for AI-powered assistants.
Sam Altman, CEO of OpenAI, mentioned there can be “a variety of various things launching this yr.” Whereas that is anticipated to incorporate Sora, the AI video device might embrace a brand new AI voice system.
What can we find out about Voice Engine?
We do not know a lot about Voice Engine or if it will likely be a product. OpenAI has not commented publicly on the matter, so all now we have are rumors and the trademark software.
Whereas Voice Engine could possibly be a brand new mannequin constructed particularly for voice functions, it is also prone to be a part of an enterprise play for OpenAI. It could possibly be constructing a high-quality voice system that might permit firms to create extra environment friendly name middle robots.
It seems rather a lot like all of the items you’d want for a completely useful and interactive AI voice assistant that may not solely deal with complicated duties, but in addition chat naturally and even take cellphone calls in your behalf.
The brand new trademark software was filed with the U.S. Patent and Trademark Workplace final week. Whereas an app does not essentially imply it should end in a product, this aligns with a broader market transferring extra towards OpenAI’s voice and route towards particular fashions.
The presentation covers the creation of software program used to create digital voice assistants, producing audio from textual content prompts, processing voice instructions, and offering voice companies.
The entire software covers the event of voice service supply, utilizing AI for textual content or speech and textual content to audio, pure language and speech processing, producing audio and speech from a message (textual content, speech, visible, picture) , voice command processing, voice recognition and creation of digital voice assistants.
That sounds rather a lot like all of the items you’d want for a completely useful, interactive AI voice assistant that may deal with complicated duties, chat naturally, and even take cellphone calls in your behalf.
The place does GPT-5 match into this?
OpenAI launched GPT-4 a yr in the past. On the time, it was an revolutionary generative AI mannequin powering ChatGPT and Microsoft Copilot.
The corporate additionally started coaching GPT-5 late final yr, sparking hypothesis about its launch date. Altman mentioned podcaster Lex Fridman“We are going to launch an incredible new mannequin this yr,” however we is not going to affirm whether or not it’s the GPT-5 or some precursor.
He additionally mentioned that “a variety of various things” can be launched within the coming months. In response to Mira Murati, CTO of OpenAI, it will embrace the Sora AI video platform.
There may be some hypothesis on social media that Sora and this new speech engine are completely different modal interfaces for GPT-5.
GPT-5 is most probably a real multimodal mannequin, able to understanding video, photos, voice, textual content, and code, in addition to producing all of these sorts of content material.
Voice Engine could possibly be a brand new assistant
Given the model description, it is also potential that Voice Engine is a brand new voice assistant, merging the broader capabilities of Siri, Alexa, or Google Assistant with the pure language and reasoning capabilities of ChatGPT.
Google has already began updating Gemini to work that means, Apple is rumored to be creating a brand new model of Siri with nice language mannequin performance, and Amazon is already testing Alexa Plus with comparable underlying expertise.
OpenAI might provide Voice Engine to energy such programs sooner or later or instead interface to ChatGPT that may run on good audio system, telephones, and even headsets.
Or it might merely be OpenAI being cautious about logos. A bid to guard GPT was rejected, so it has now filed trademark functions for GPT-5, 6, and even GPT-7. The latter consists of producing music, changing textual content and knowledge into code, and writing code from scratch.
- Google’s new VLOGGER AI helps you to create a sensible avatar from only a photograph
- Apple might deliver Google Gemini to the iPhone for AI duties
- I gave Claude, ChatGPT and Gemini a photograph of the elements to see which one got here up with one of the best recipe.
