Soon, ChatGPT could become an intelligent voice assistant thanks to Voice Engine. Siri, Alexa and Google Assistant would then have to worry about.

Sam Altman made some interesting revelations during his latest appearance on Lex Fridman’s podcast. The boss of OpenAI notably announced the future release, this year, of a “ amazing new model “. It would be about Voice Engineof which the Californian start-up recently made a trademark registration application.

ChatGPT has shaken up the communication sector online search. The popular generative conversational agent would now attack that of intelligent voice assistants.

OpenAI filed an application to make Voice Engine her brand. Remember that it is the United States Patent and Trademark Office (USPTO) which handles these requests.

The submission of the file took place on Monday March 18, the day after Altman’s interview with Fridman. During this interview, the CEO indicated that the Californian start-up had “ lots of other important things to come out ” Before GPT-5.

But what do we know about Voice Engine?

The trademark registration document obviously details the main fields of action of this mysterious vocal engine. These are:

  • Designing intelligent voice assistants
  • Speech and voice recognition
  • Processing voice commands and conversion between speech and text
  • Automatic recognition and generation voice and speech
  • Generating speech or sounds using prompts in natural language (text, speech, images and videos)
  • Generation of voice and audio signals with prompts
  • Natural language and speech processing usingmachine learning
  • Multilingual speech recognition, translation and transcription
  • Use of artificial intelligence for automatic text-to-speech and text-to-audio conversion
  • Development of voice services

This is all well and good, but it is not certain that the USPTO will validate the trademark registration. Remember that last month, OpenAI did not win its case for make GPT its brand.

On the other hand, the Californian start-up does not yet offer an intelligent voice assistant to its users. However, it provides them with a API to convert speech to text.

Another reminder, OpenAI has a speech recognition model called Whisper. This one is much less popular than the GPT models of the start-up.

GPT-5: Not for now!

Before Voice Engine, OpenAI applied for a trademark registration for GPT-6 And GPT-7. These are the future major language models (LLM) of the start-up.

There conversation simulation is one of the major new features of GPT-6, while GPT-7 will allow ChatGPT to generate music.

Sam Altman also mentioned the case of GPT-5. Users who are hoping for the successor of GPT-4 and of GPT-4 Turbo will have to be patient.

We have a lot of other important things to get out first », Says the boss of OpenAI. Voice Engine would be one of those important things.

    Share the article:


Our blog is powered by readers. When you purchase through links on our site, we may earn an affiliate commission.





Source link

Leave a Reply

Your email address will not be published. Required fields are marked *