OpenAI recently introduced the research preview of DALL-E 3, their latest image generation AI. This exciting development will soon be accessible to ChatGPT Plus and Enterprise users. The integration of DALL-E 3 with ChatGPT simplifies prompt creation with the assistance of the chatbot.
DALL-E 3's capacity to comprehend image inputs relies on GPT-4 Vision (GPT-4V), a multimodal version of the underlying GPT model. Additionally, the voice feature leverages OpenAI's Whisper automatic speech recognition (ASR) model for processing user voice inputs. Furthermore, a new text-to-speech (TTS) model allows ChatGPT to convert its text responses into one of five user-selectable voices.
OpenAI is taking a gradual approach to deploy these features, prioritizing safety. They've conducted beta testing and 'red teaming' exercises to identify and mitigate potential risks. OpenAI's commitment to ensuring secure and efficient AI usage is at the core of these developments.
Comments