In the past two weeks, OpenAI has introduced new features in ChatGPT, expanding user interaction and capabilities. The following updates are noteworthy:

1. Bing and DALL-E 3 Integrations:
   • ChatGPT now facilitates web searches via Bing and image generation through DALL-E 3. The Bing integration allows users to access current information published after ChatGPT’s last training dataset update. When ChatGPT encounters a question for which there’s no suitable answer in its training dataset, it can now look up relevant information on the web using Bing. Initially introduced in May, the Bing integration was briefly disabled after users discovered the chatbot could retrieve paywalled website content without permission. OpenAI re-enabled this feature last month but kept it in beta until this week​¹​.
   • The new image generation feature enables the creation of images by entering natural language prompts into the ChatGPT interface. This feature is powered by DALL-E 3, the latest addition to the DALL-E series of AI image generators launched by OpenAI in 2021. According to OpenAI, DALL-E 3 generates higher-quality images than its predecessors and can interpret user requests more accurately​¹​.
2. GPT-4V (Vision Capability):
   • Another significant update is the introduction of a groundbreaking vision capability named GPT-4V in ChatGPT, allowing users to upload and analyze images within ChatGPT​²​.
3. Speech and Hearing Capabilities:
   • ChatGPT has also acquired the ability to interact with users through speech and images, significantly enhancing the user experience​³​.

These updates signify OpenAI’s continuous progress in developing ChatGPT, and represent substantial improvements concerning interactivity and user experience.