ChatGPT's major update, A rapid improvement in multimodal capabilities

September 25th, ChatGPT launched major updates to its platform with the introduction of image and voice capabilities. The voice feature is available for both iOS and Android mobile platforms, while the image feature is available on all platforms.

The new features are expected to be available to Plus and Enterprise users within two weeks and will be subsequently released to developers and other user groups.

AI art generated in DALL-E 3

Based on the demonstrations provided on the OpenAI website, we believe that ChatGPT has significantly improved its image recognition and analysis capabilities with the addition of these features. The voice feature also provides a more convenient and faster way to interact with the platform. Overall, the two new features significantly enhance the user experience and expand the potential use cases of ChatGPT. We anticipate that OpenAI's subsequent release of related APIs will facilitate the integration of image and voice-related AI capabilities into various applications, further expanding the possibilities of AI applications.

DALL·E 3, which is built natively into ChatGPT, significantly enhances the platform's image capabilities. On September 20th, OpenAI released a preview version of its large-scale generative model DALL·E 3 for a small group of test users. The model is expected to be available to Plus and Enterprise users in early October and later to research laboratories and APIs in the fall. Compared to DALL·E 2 and Midjourney, we believe that DALL·E 3 significantly improves image quality and better understands and strictly adheres to user instructions. The native integration of DALL·E 3 into ChatGPT will greatly enhance the platform's user experience and expand its potential use cases. ChatGPT can also serve as an inspiration assistant and prompt rewriter for users using DALL·E 3 to create images, significantly improving the effectiveness of generative art. The two technologies will mutually reinforce each other, creating a positive impact.

$Microsoft(MSFT)$ $Alphabet(GOOG)$ $Alphabet(GOOGL)$