- India
- International
Google firmly believes that Artificial Intelligence will help everyone, from professionals to students to parents to developers, to reach their full potential. The company has been consistently making advancements in its language research and its models are being integrated into a wide range of products and services.
In March this year, when the world was witnessing groundbreaking innovations in AI, Google unveiled its conversational generative artificial intelligence chatbot, Bard. Ever since its launch, Google has been working relentlessly to make its chatbot the most efficient AI assistant. Recently, Google moved Bard to PaLM 2, a far more capable language model that has led to many recent improvements.
This year at the Google I/O event, the company introduced a slew of upgrades to Bard. Google acknowledged that it has made progress with Bard and the new features showcase the company’s broad vision for the future of Bard.
Most people have been using Bard to enhance their productivity and make their lives easier. From a filmmaker who needs help in delivering a pitch to a busy parent unable to plan their holiday family dinner or an entrepreneur who needs support with brainstorming for ideas, Bard could be a great collaborative partner.
At I/O, Google announced updates to Bard including wider availability, multimodality, apps and extensions, and new coding features.
Bard will now be open to 180 countries and Google is doing so by removing the waitlist and allowing more people to try it in English. And, talking about English, Bard will also be offered in Japanese and Korean. These are the first two languages beyond English that are being launched by Google as a mechanism to responsibly roll out the product, gain feedback and use the information to help it roll out 40 more languages in the future. Since large language models are still in the nascent stage with numerous limitations, Google asserts that it will roll out all the upgrades in adherence to its AI principles.
Perhaps, this is the biggest upgrade to Google Bard. The chatbot was launched with just text, but now Google is endowing it with richer responses. Users will be able to see images directly from Google Image Search in their responses. For instance, if a user is travelling to Mumbai and if they use the prompt ‘Where should I travel in Mumbai?’ Bard will include visuals in line with the text in its response.
Google will also allow users to prompt Bard with images using Google Lens. A user who is stuck with art supplies and does not know how to use them for craft, can simply click a picture and ask Bard what can be done with it. Google Bard will instantly come up with some interesting craft ideas.
Along with multi-modality, Google will also soon bring extensions to Bard and introduce some of its own apps and services directly to the Bard experience. The company has highlighted that users will be in control of their privacy and security.
Most importantly, Bard will also be able to harness various services from the Internet with extensions from some of the most incredible partners. The company will soon integrate Adobe Firefly directly into Bard to help users generate brand-new images.
Besides, Google is also working towards connecting Bard with helpful Google apps and more partners including, OpenTable, Kayak, ZipRecruiter, Instacart, Wolfram and Khan Academy.
Google said that it was really important for it to build Bard alongside its users and that their feedback was key to making the product better. As part of this, starting next week Google will ensure that code citations are even more precise and will show the source of specific blocks of code. The company said that citations will also apply to narrative content from across the web.
Additionally, Google will also add an Export button next week. This is considering the feedback from developers who said that they enjoyed some of the explanations and generation that they saw and wanted to export the code lab, specifically from the Python lab. Developers will be able to export into the Restlet IDE as well to build and reiterate their ideas.
Google has also introduced a dark theme. This was specifically since many developers sought a theme that was a ‘little easier on the eyes’. To add to the ease of using Bard to help draft emails and documents, Google has introduced export actions that make it easy to move from an idea in Bard to a draft in Gmail or in Docs.
According to Jack Krafcik, Senior Product Director at Google, Bard was launched in the US and the UK on a lightweight large language model to get feedback and iterate quickly. And, since the launch the company has made rapid progress. In the last few months, improvements enhance the maths, logic, reasoning skills, and coding capabilities of Bard which can not only generate code but also debug and explain code now. As of now, Bard understands 20 programming languages, and this capability has made it among the most favourite AI tools to engage and collaborate with.