Google Unveils Whisk, a New AI Tool for Generating Images with Visual Prompts

Introducing Whisk: A New Way to Generate Images with Visual Prompts

Google has announced a new AI tool called Whisk, which allows users to generate images using other images as prompts instead of requiring a long text prompt. With Whisk, you can provide images to suggest the subject, scene, and style of your AI-generated image, and you can prompt Whisk with multiple images for each of those three things. If you don’t have images on hand, you can click a dice icon to have Google fill in some images for the prompts (although those images also appear to be AI-generated).

How Whisk Works

Whisk will generate images and a text prompt for each image. You can favorite or download the image if you’re happy with the results, or you can refine an image by entering more text into the text box or clicking the image and editing the text prompt.

A Hands-On Experience with Whisk

I had the opportunity to try out Whisk and was able to generate a series of images by iterating on the prompts. I started by clicking the dice icon to generate a subject, scene, and style, and then swapped out the auto-generated scene by entering a text prompt. Whisk created the first two images, which I then refined by asking Whisk to add some steam around the subject (because it’s a fire being in water), resulting in the next two images.

Google’s Vision for Whisk

In a blog post, Google stresses that Whisk is designed for "rapid visual exploration, not pixel-perfect edits." The company also notes that Whisk may "miss the mark," which is why it lets you edit the underlying prompts. Google is confident that Whisk will be a valuable tool for creatives, designers, and anyone looking to generate high-quality images quickly.

Conclusion

Whisk is an exciting new tool that has the potential to revolutionize the way we generate images. With its ability to use visual prompts, Whisk offers a more intuitive and creative way to generate images, and its rapid iteration capabilities make it a great tool for quick visual exploration.

Frequently Asked Questions

Q: What is Whisk?
A: Whisk is a new AI tool from Google that allows users to generate images using other images as prompts instead of requiring a long text prompt.

Q: How does Whisk work?
A: Whisk generates images and a text prompt for each image, and users can refine the prompts by entering more text or editing the text prompt.

Q: What is the purpose of Whisk?
A: Whisk is designed for "rapid visual exploration, not pixel-perfect edits," and is intended to be a valuable tool for creatives, designers, and anyone looking to generate high-quality images quickly.

Q: Is Whisk available for use?
A: Yes, Whisk is available for use, and users can try it out to generate their own images.

ChatGPT’s AI Search Engine Rolls Out to All Users

OpenAI has announced that ChatGPT’s AI search engine will be rolling out to all users starting today. As part of its newest “12 Days of Ship-Mas” livestream, the company also revealed an “optimized” version of the feature on mobile and the ability to search with advanced voice mode.

Improvements for Mobile Search

One of the key improvements for search on mobile devices is the ability to display results in a more traditional search engine format. For example, when searching for a specific location, such as restaurants or local attractions, ChatGPT will display a list of results with accompanying images, ratings, and hours. Tapping on a location will pull up more information about the spot, and users can also view a map with directions from directly within the app.

Enhanced Search Features

Another feature aimed at improving the search experience is the ability to surface links to websites quickly when searching for specific types of sites, such as “hotel booking websites”. Instead of generating a response right away, ChatGPT will display links to relevant websites before providing more information about each option. Additionally, the advanced voice mode will be available for paid users, allowing for automatic provision of up-to-date information from the web.

Conclusion

The rollout of ChatGPT’s AI search engine to all users marks an exciting milestone in the development of this revolutionary technology. With its ability to provide faster and more accurate search results, improved mobile functionality, and enhanced features, ChatGPT is poised to change the way we interact with search engines forever.

Frequently Asked Questions

Q: When will the AI search engine be available to all users?

A: The AI search engine is available starting today.

Q: What are the new features being added to mobile search?

A: The new features include the ability to display results in a more traditional search engine format, with images, ratings, and hours, as well as the ability to view a map with directions from directly within the app.

Q: What is advanced voice mode, and how does it work?

A: Advanced voice mode is a feature that allows users to search with their voice, and it is only available to paid users. It automatically provides up-to-date information from the web.

Q: Is the advanced voice mode the same as voice search on other search engines?

A: No, ChatGPT’s advanced voice mode is a unique feature that uses AI to provide more accurate and relevant results.

YouTube to Allow Creators to License Videos for AI Training

YouTube is rolling out a new feature that will allow creators to license their videos for use in training artificial intelligence (AI) models. The default setting for this feature is off, meaning that creators don’t have to do anything if they don’t want to allow third-party companies to use their videos for AI training. However, creators who do want to allow this will be given the option to do so.

How it Works

According to a support post by a TeamYouTube staffer named Rob, creators will be able to choose from a list of third-party companies that can train on their videos. They will also have the option to allow all third-party companies to train on their videos. The initial list of companies includes AI21 Labs, Adobe, Amazon, Anthropic, Apple, ByteDance, Cohere, IBM, Meta, Microsoft, Nvidia, OpenAI, Perplexity, Pika Labs, Runway, Stability AI, and xAI.

Google’s Perspective

A YouTube spokesperson, Jack Malon, said that the list of companies was chosen because they are building generative AI models and are likely sensible choices for a potential partnership with creators.

What This Means

This announcement comes as reports of AI models from big companies, including OpenAI, Apple, and Anthropic, being trained on content and datasets scraped from YouTube have been circulating. Google itself already uses YouTube data to help train its AI tools.

Conclusion

The new feature is a significant development in the world of AI and content creation. It provides creators with more control over how their content is used and potentially allows for new forms of collaboration between creators and third-party companies.

FAQs

Q: What is the default setting for this feature?
A: The default setting is off, meaning that creators don’t have to do anything if they don’t want to allow third-party companies to use their videos for AI training.

Q: How can I opt-in to allow third-party companies to train on my videos?
A: Creators will be able to choose from a list of third-party companies that can train on their videos or allow all third-party companies to train on their videos.

Q: How was the list of companies chosen?
A: The list of companies was chosen because they are building generative AI models and are likely sensible choices for a potential partnership with creators.

Q: How does Google use YouTube data?
A: Google uses YouTube data to help train its AI tools, consistent with the terms that creators agree to.