News and Tips

ChatGPT Evolves: Now With Voice and Image Features

ChatGPT news
If you already have the Chat GPT in everyday life, prepare yourself for an even more in-depth interaction! The introduction of these new features promises to transform the way we communicate with this virtual assistant. Now, it will be possible not only to talk to him, but also show images, further enriching the dialogue.
Those Updates open doors to a variety of practical uses. Suppose you are traveling and decide to share a photo of a place with your Chat GPT. You will be able to discuss it in real time. At home, you can take photos of the items in your fridge and brainstorm dish ideas. You can even receive detailed recipes.
Soon, the OpenAI plans to make these innovations available to users Plus and Enterprise. It is worth noting that the voice feature will be enabled exclusively for mobile applications, while the image functionality can be explored on all available platforms.
Stay tuned to try out these new ways of interacting and discover everything the revamped ChatGPT has to offer!

Chatting with ChatGPT

Previously, ChatGPT could only listen to you, but with recent updates, it is also able to respond to you audibly!
The innovative voice function allows the user to interact with ChatGPT, which can now respond through five different synthesized voices of their choice.
To take advantage of this feature, go to Settings → New features in the mobile app and select the option for voice conversations. Then, click on the headphone icon located in the top right corner of the home screen and select your favorite voice from the five available.
According to OpenAI, voice functionality is powered by a sophisticated voice model text to speech conversion, developed based on samples from voice actors. For speech recognition, Whisper, OpenAI's open source speech recognition system, is employed.

Present Images to ChatGPT to Contextualize Conversations

You now have the ability to enrich your interactions with ChatGPT by displaying one or multiple images, providing visual context and directing the conversation.
As an illustration, it is possible to send an image of a damaged bicycle and ask the chatbot to identify the fault and propose solutions. On mobile devices, a drawing feature makes it possible to highlight or indicate specific areas in the image.
The new visuals are incorporated thanks to a multimodal version of the templates GPT-3.5 and GPT-4, adapted to interpret visual information. It is worth noting that, before its implementation, OpenAI carried out extensive testing on the visual resources to ensure security.

Cautious Implementation with an Emphasis on Security

OpenAI is following a phased strategy for introducing these new features.
Innovative voice technology not only provides creative uses, it also presents challenges, such as imitating well-known personalities. In order to minimize these risks, voice functionality is, for now, restricted to conversations.
As for images, OpenAI has limited the ChatGPT's ability to examine people in photographs directly and recommends caution in high-risk applications without due verification.
The recent additions of voice and image capabilities to ChatGPT provide more intuitive interaction with the chat system. IA. Users Plus and Enterprise will have the opportunity to explore these features in the coming weeks.
Artificial Intelligence