OpenAI has rolled out new voice and image capabilities for its popular AI-powered chatbot, ChatGPT. These new capabilities allow users to have more natural conversations with ChatGPT by speaking to it and showing it images.
Voice capabilities
The voice capabilities in ChatGPT are powered by a text-to-speech model. This model can generate realistic human-sounding speech from text. Users can speak to ChatGPT by pressing a microphone button on the screen. ChatGPT will then transcribe the user’s speech to text and respond with a text or voice response.
The voice capabilities in ChatGPT are still under development, but they are already very impressive. ChatGPT can understand and respond to a wide range of voice prompts. For example, users can ask ChatGPT to generate different creative text formats of text content, like poems, code, scripts, musical pieces, email, letters, etc., or to answer their questions in a comprehensive and informative way, even if they are open ended, challenging, or strange.
Image capabilities
The image capabilities in ChatGPT are powered by multimodal GPT-3.5 and GPT-4 models. These models can understand the content of images and generate text that is relevant to the image. Users can show ChatGPT an image by uploading it from their device or by taking a photo with their camera. ChatGPT will then analyze the image and generate a text response.
The image capabilities in ChatGPT are also still under development, but they are already very useful. ChatGPT can identify and describe objects in images, as well as understand and respond to complex image prompts. For example, users can ask ChatGPT to describe the contents of an image, to identify and label objects in an image, or to generate a creative text format of text content based on an image.
The new voice and image capabilities in ChatGPT are a significant improvement. They make it easier to have more natural conversations with ChatGPT and make ChatGPT more versatile and useful. These new capabilities are still under development, but they are already very impressive. I am excited to see how OpenAI continues to develop these capabilities in the future.
