Google’s Chatbot Bard Released to Over 180 Countries
In a groundbreaking announcement at Google I/O, Sissie Hsiao, vice president at Google, declared that the company’s generative AI chatbot, Bard, will now be widely available.
The waitlist restrictions that previously limited access to this innovative tool have been removed, making it accessible for users across 180 countries and territories in English, with plans to extend this reach further soon.
“We want to get Bard into more people’s hands so they can try it out and share their feedback with us,” Hsiao wrote in a blog post outlining the change.
The broader release is a part of Google’s strategy to continually develop and refine the chatbot with input from a more extensive user base.
John Krawczyk, senior product director at Google, also revealed that versions of the chatbot in Korean and Japanese are in the pipeline. Google plans to expand Bard’s linguistic repertoire to the top 40 languages soon after the conference, he confirmed.
The company has been cautious in its approach to language expansion, rolling out different languages separately as a safety measure. This is in line with Google’s commitment to responsible development, which was a recurring theme at the press event. Bard, referred to as an “experiment,” is poised to transform our interaction with digital interfaces.
Bard’s Multimodal Future
In a significant upgrade, Google plans to add multimodal content to Bard, enabling it to provide answers in forms beyond just text.
Users will soon be able to ask queries like, “What’s a fast, balanced, vegetarian meal for me to make?” and receive a response complemented with rich visuals that enhance the understanding of the subject matter.
This initial visual support will encompass pictures, with prospects to include maps, charts, and other items in the future.
Bard Vs. Rivals
The recent enhancements to Bard come two months after its limited release for users in the US and UK. These upgrades, which include the use of Google’s new PaLM 2 language model, are set to improve Bard’s overall performance and usability.
Despite these improvements, comparisons with rivals such as OpenAI’s ChatGPT and Microsoft’s new Bing chatbot have highlighted some shortcomings, including factual errors in Bard’s initial public demo.
Practical Features and Coding Enhancements
Among the new features in Bard are practical tools like an export button for direct transfer of text to Gmail or Google Docs.
The upgraded Bard is designed to efficiently handle coding queries, including debugging and explaining code snippets in over 20 languages.
This includes the integration with browser-based IDE Replit (starting with Python queries) and improved citation for code that not only provides sources but also explains the snippets.
Bard is also gaining visual capabilities, including the ability to analyse images, offer images in query results, and generate visuals using AI, a feature powered by Adobe’s Firefly software. This will allow Bard to produce visual results akin to Google’s image search in response to location-based queries.
Google also plans to integrate Bard with third-party web services like Instacart and OpenTable.
This mirrors the moves of rivals Microsoft and OpenAI, and adds to the suite of features designed to make Bard a more useful and versatile tool.
The integration with Adobe’s Firefly is particularly significant, as it allows Bard to generate images. Adobe has marketed Firefly based on the ethical nature of its training data, an aspect that has been a point of controversy for other AI image tools.
Bard’s Future: More than Just Search
While Google has emphasised that Bard is not a replacement for its search engine, the bot’s enhanced features make it a viable alternative for many search-related tasks. However, the true purpose of Bard remains somewhat elusive.
As Google continues to infuse AI across its various domains and initiates changes to its search engine, Bard seems to be morphing into an experimental playground for Google’s AI ambitions.
An Array of Features and Integrations
The removal of the waitlist at Google I/O means that users in over 180 countries can now access Bard immediately, provided their language and country are supported.
Google will gradually expand this access to more languages, countries, and regions over time.
Bard is also being enriched with a more visual interface, Google Lens support, dark mode, and a range of integrations and features.
Google’s search and knowledge graph data have been harnessed to make Bard more visual and useful, with responses supplemented by rich visuals for an enhanced user experience.
The integration with Google Maps means Bard can now display maps directly within its answer box.
Google Lens, a feature that allows Bard to analyse photos and craft creative captions, is another exciting addition.
Google has also improved Bard’s coding assistance capabilities. Precise citations for blocks of code or other content will be available starting next week. Users can simply click the annotation and Bard will underline the response parts linked to the source. An “Export” button has also been added, allowing users to export and run code with Google’s partner Replit, starting with Python.
The chatbot’s appearance can now be switched between a light background with dark text to a dark background with light text, thanks to the newly added dark mode.
Google has also included one-click options to export content, including formatting, directly into Google Docs and Gmail.
The integration with Adobe Firefly further expands Bard’s capabilities. Users can now easily turn their creative ideas into high-quality images, which can then be further edited or added to their designs.
As Google steps up to the plate to compete with other AI chatbots, Bard’s enhancement and expansion demonstrate the tech giant’s ambition to lead in this innovative technology space.
