Bard is Back: Google’s Answer to ChatGPT Gets An Upgrade

The launch of Google's AI chatbot, Bard, felt very rushed. Microsoft is a big investor in Open AI and has leveraged it to improve Bing, gaining a 16% increase in traffic, according to SimilarWeb data, since the integration.

Google has enjoyed a lengthy period of dominance in the Search space. The speed and strength of the challenge from Open AI & Bing would definitely have set off a code red at Google. AI is powerful enough to completely disrupt Google's main source of income. This seemed to force Google's hand and they released a demo version of Bard... a bit too soon.

An Awkward Start

In its initial public demonstration, a GIF released by Google showed Bard answering a query about the James Webb Space Telescope. Yet, one of the points it made stated that the telescope had captured the first-ever images of a planet outside our solar system. Twitter had plenty to say of course:

Quite embarrassing.

But Bard is Back!

Fast forward to the recent Google I/O conference and Bard is back, stronger and smarter and way more polished. Google announced Bard's availability in over 180 countries, touting support for Japanese and Korean languages, visual responses to queries, integration with Google services, and powerful extensions enhancing its capabilities.

The revamp includes a significant upgrade in the form of Google's new PaLM 2, a family of large language models (LLMs) known for their advanced math and reasoning skills and superior coding capabilities. This upgrade marks a transition from Bard's previous reliance on Google's LaMDA AI model.

A Host of New Features

Google has ambitious plans to integrate Google Lens with Bard, allowing users to involve photos and images in their prompts. The integration means you could upload a picture of your pets and ask Bard to generate a humorous caption, for instance. Bard will not only detect the breeds but also deliver amusing captions on-demand.

Developers will appreciate several key improvements in Bard, such as an export button for code execution in either Colab or Replit. Bard will also provide code citations to encourage transparency and traceability. This is a really important feature, given how sometimes AI models seem to like to make things up. To top it off, the chatbot will now support a dark theme, which notably garnered the loudest cheer at the I/O keynote.

Expanding Bard’s Horizons

Google plans to enhance Bard with extensions, similar to ChatGPT plugins, allowing it to access external resources. For instance, Google showcased how Bard could generate images of unicorns for a birthday party using an Adobe Firefly extension.

Personally, I'm looking forward to being able to generate visuals outside of a Midjourney Discord server!

Partnerships with service providers like Kayak, OpenTable, ZipRecruiter, Instacart, Wolfram, and Khan Academy are also in the works to build extensions for Bard. Further, Google will integrate features from existing apps like Docs, Drive, Gmail, and Maps into Bard's user experience over time. These efforts align with Google's vision of integrating generative AI into all its products.

The Gmail integration is probably the one that will have most impact in the short term, to most people. Think ChatGPT style 'write me a highly persuasive email to my boss to ask for a raise' but directly within Gmail.

Bard Is Open for Business

Google Bard can now be accessed by anyone in the 180 supported countries with a Google account:

Bard may have had a shaky start, but its rapid transformation serves as a testament to Google's ambition in the AI space.... or maybe just highlights how pretty much every organisation has no choice but to build on AI or die out.