surprised to see the innovative and unique ways people have connected with Bard, our project that enables you to work with generative AI, less than two months after we introduced it. (I, for one, have learned some pretty entertaining methods for teaching fractions to my 7-year-old!)
Since we launched Bard, originally in the U.S. and the U.K., we’ve received a lot of feedback and have swiftly modified it to improve your experience. We recently switched Bard to PaLM 2, a far more powerful big language model, which has allowed many of our latest advancements, including sophisticated mathematical and logical thinking abilities as well as coding skills. Coding has rapidly risen in popularity over the last several weeks as a result of Bard.
But, this initial impetus is just the start. In addition to revealing more about our future plans, we are presenting new methods for you to work with Bard now.
introducing Bard to wider customers
We want to get Bard into more people’s hands so they can try it out and give us their feedback as we continue to create further enhancements and add new features. As a result, we are deleting the queue today and making Bard available to more than 180 nations and territories.
Not only that, but Bard is now accessible in Korean and Japanese, and we expect to offer 40 languages by the end of the year. Large language models are still a developing technology with acknowledged drawbacks, as we have said from the outset. Hence, as we continue to grow, we’ll make sure to uphold our AI Principles and keep up our high standards for quality and regional specifics.
Visualizing your interactions with Bard
Soon, both Bard’s replies and your suggestions will include additional visual elements. You’ll be able to ask it questions like, “What are some must-see attractions in New Orleans?” and receive answers that are both text-based and packed with images so you can get a greater feel of what you’re looking at.
Together with text, you will be able to use pictures in your own prompts, which will provide you with new opportunities to exercise your imagination and creativity. We’re bringing the power of Google Lens directly to Bard to make this happen. Let’s imagine you want to use a picture of your pets to have fun. Upload it and tell Bard to “create a witty caption about these two,” then click “Upload.” Within seconds, Bard will evaluate the image, identify the breeds of the dogs, and come up with a few clever captions.
introducing export capabilities and code updates
We must create Bard with users since their input is essential to improving it. We’re implementing developers’ suggestions into a few significant code enhancements as part of that endeavor, such as the following:
- Source citations: We’ll start becoming more specific with citations the following week. Just click the annotation to highlight any code or other material that Bard includes in the answer and to connect to the relevant source.
- Dark theme: We are introducing the Dark theme today because we believe it will make using Bard much easier on your eyes. This is another feature that developers have requested.
- “Export” Button: Since the export to Colab feature is well-liked by developers, we’re adding the ability to export and run code with our partner Replit, beginning with Python, in the near future.
Also, we’re introducing two new Export Actions today to make it simple to import Bard’s replies directly into Gmail and Documents. This is because users often ask Bard for a head start when composing emails and other documents. Let’s pretend, for illustration, that, like myself, you are an avid pickleball player. You may ask Bard to create an email announcement for your new pickleball league that summarizes the regulations and emphasizes how the sport is open to players of all ages and skill levels. To make any last-minute adjustments before starting your pickleball league, just click the “draft in Gmail” option.
Providing the services you love in Bard
By incorporating the features of Google applications and services you may currently use, like Docs, Drive, Gmail, Maps, and others, directly into the Bard experience, we’ll soon bring new methods to spark your creativity and curiosity. In addition, you will always be in charge of how you utilize these tools and extensions in terms of privacy settings.
With extensions from third parties, Bard will also be able to access a variety of services from around the internet, enabling you to take advantage of features never previously available. We’ll integrate Adobe Firefly, a collection of artistic generative AI models, into Bard in the next few months so you can quickly and simply convert your own creative concepts into high-quality pictures that you can subsequently modify or include in your designs in Adobe Express.
Imagine I’m organizing a unicorn-themed birthday party for my 7-year-old. I want to include a cute picture with the invites. All I have to do is tell Bard to “make a picture of a unicorn and a cake at a kids party,” and it will do it quickly while upholding Adobe’s strict standards for morality and quality.
We’re working to integrate Bard with practical Google products and many more partners, like Kayak, OpenTable, ZipRecruiter, Instacart, Wolfram, and Khan Academy, because we want Bard to be a home for your creativity, productivity, and curiosity.
Bard has a lot of work ahead of it, linking Google products and incredible online services to enable you to perform and create whatever you can dream of via a seamless partnership with our most powerful big language models. The possibilities are endless when we mix human creativity with Bard’s generative AI powers. We’re eager to see what you do with it.
