👨🏻‍🎨ChatGPT’s New Canvas

OpenAI launches new ‘canvas’ interface for ChatGPT, designed for writing and coding tasks

Key Points 

• OpenAI's new "canvas" interface allows users to work side-by-side with ChatGPT, making it easier to adjust and refine generated text or code.

• The interface opens an editing window next to the chat box, providing shortcuts for inline edits, grammar checks, and coding tasks.

• The canvas interface is currently in beta and can be accessed by selecting “GPT-4o with canvas” in ChatGPT. 

👨🏻‍💻News - OpenAI has introduced a new "canvas" interface for ChatGPT that lets you work side-by-side with the chatbot, adjusting parts of the text or code it generates.

🤹🏻‍♂️An overview - Canvas opens an editing window next to the chat box, allowing you to manually tweak text or code and highlight parts for more focused feedback. It includes a menu of shortcuts for inline edits, grammar and clarity checks, and adjusting text length and reading level.

For coding, there are shortcuts for debugging, adding logs and comments, and translating code. The concept is similar to the "Artifacts" feature that Anthropic launched for Claude back in August.

🌟Standout features - Canvas takes away the need to keep entering new prompts to refine ChatGPT's responses. It makes everything simpler and gives you more control, with features like a back button to restore older versions of your work. OpenAI is also fine-tuning when Canvas activates, ensuring it only kicks in when needed, so it won’t disrupt your flow.

▶️How to get started - The canvas interface is currently in beta and you can access it by selecting “GPT-4o with canvas” from the model picker in ChatGPT. It’s being rolled out globally to Plus and Teams users, with Enterprise and Edu users set to get access next week. Once it’s out of beta, free ChatGPT users will also be able to use it. 

Google adds AI video search feature to Lens, revamps AI Overviews

Key Points 

• Google Lens now lets you take videos for searches and ask questions with your voice, generating helpful AI Overviews and relevant results.

• Google is also rolling out AI-organized search pages and better formatting for AI Overviews, which now include ads for product suggestions. 

🎥News - Google Lens now allows users to take videos for searches and ask questions using voice commands. This new feature will generate an AI Overview and relevant search results based on the video and the user's inquiry. 

For example, if you’re at an aquarium and you see some fish, you can point your phone at the exhibit, start recording, and ask, “Why are they swimming together?” Google Lens will then use its Gemini AI model to give you an answer. This feature is rolling out in Search Labs for both Android and iOS.

🤔How exactly does it work though? The new video search feature captures video as a series of image frames, using the same computer vision methods from Google Lens. What’s neat is that it uses a custom Gemini model that can understand sequences of frames, which helps it provide answers based on information online.

Right now, it can’t identify sounds in videos—like if you want to know what bird you're hearing—but Google is apparently working on that.

💁🏻‍♀️See also - 

• Google is rolling out ads in AI Overviews, meaning you'll see product suggestions alongside tips when you search for something like how to remove a grass stain. These will appear under a “sponsored” header.

• The formatting of AI Overviews is improving, with cited webpages now displayed more prominently on the right side of the summary.

• Google is introducing AI-organized search pages, which present a customized results page with relevant information instead of just a list of links. This feature is currently available only on mobile in the US for searches related to recipes and meal ideas.