Google is outfitting Bard(opens in a new tab) with some meaningful upgrades with an assist from Adobe.
As anticipated, the tech company announced major AI features related to its AI chatbot. Bard is now powered by PaLM 2, which was also announced today, and new updates include tools for coding, multimodal capabilities, and integration with other Google tools like Docs and Sheets.
The AI chatbot now supports multimodal search, meaning it's capable of generating images from a text prompt. Bard responses will include visuals in addition to text, and it can generate images for you based on text. Image generation is supported by Adobe Firefly(opens in a new tab) which will share images directly in the Bard chat.
With the multimodal update, you can also share an image in the chat and get a text response. There's also a new export button for sharing Bard responses to Gmail and Docs.
Bard is now integrated with Google tools like Docs and Sheets, which means it can generate content and tables, and automatically add them to these apps. For coders, Bard also added 20 new programming languages and citations for understanding code generation. The export button can also export and run code starting with Python. The biggest applause though, came from Bard's introduction of "dark mode."
Bard was launched in February as Google's response to OpenAI's ChatGPT and Microsoft's Bing which is powered by GPT-4. But due to inaccuracies and less-refined responses, it's been considered a subpar model compared to its rivals. Last week, Microsoft announced multimodal features for Bing and its browser Edge. Now Google is leveling up to bring similar capabilities to its own AI chatbot.
Google has touted Bard as a tool for programmers, and today it doubled down on that with new coding, math, and logic capabilities thanks to PaLM 2, Google's more advanced LLM which was also announced today.
As of today, the waitlist for Bard is removed, so users can test it out for themselves.
