OpenAI has announced major upgrades to its popular chatbot, ChatGPT, and to Spotify, which will use AI voice features. ChatGPT will now allow users to use voice commands and make queries via images. These options will first be available to subscribers of the Plus and Enterprise plans, while other users will gain access a little later.

To enable these new functionalities on ChatGPT, users should go to settings and activate voice chats. OpenAI has developed five different chatbot voices using professional actors to give users a variety of options.

OpenAI uses a new text-to-speech model that can generate a convincing human-like voice from text or short audio samples. Also, OpenAI uses its speech recognition system, known as Whisper, to recognize and convert the user's spoken words into text.

Apart from voice features, ChatGPT will now also support image recognition functionality. Users will be able to show the chatbot images and ask questions or seek answers based on those images. This includes math problems and queries like "what's in the picture" or "how to make a dish with these ingredients".

It's important to note that OpenAI has set limits on how ChatGPT can analyze images shared by users to ensure that individuals' privacy is respected. Also, the bot is the most effective in understanding English on images.

Spotify is also introducing AI voice features

Meanwhile, Spotify announced a partnership with OpenAI to use their AI voice technology. Spotify will launch a pilot tool called "Voice Translation for podcasters" that will allow podcasts to be translated into different languages using the voices of native speakers. The tool will retain the speech characteristics of the original speakers and will first be available for select shows in English with possible expansion to other languages.

These innovations promise exciting new possibilities for ChatGPT and Spotify users, opening the door to voice communication and better understanding of images and language.