Google has built a new model for Bard which it is calling the most capable iteration of the AI yet.
Google provided an update on the new version of Bard which it calls “more intuitive, imaginative and responsive than ever before,” offering greater levels of quality and accuracy in the chatbot’s responses.
A whole bunch of new features have been brought into the mix for Bard, and that starts with support for 40+ languages, and some tight integration with existing Google products elsewhere.
That includes giving Bard the ability to get its hooks into your emails in Gmail, and data in Google Drive and Docs, meaning you can get the AI to find info across your various files, or indeed summarize a piece of content if needed.
Bard will also be able to pull data in real-time as needed from Google Maps, Google’s travel features (hotels and flights), and YouTube, all of which will be extensions that are enabled by default (you can disable them if you wish, but they’re switched on by default in the new Bard).
Another big move here is the ability to check Bard’s answers. Not too sure about any given response from the AI? A ‘Google It’ button can be clicked to bring up additional info around any query, which is drawn from Google search (where supported), so you can check for yourself to see if there’s any doubt, or difference of opinion, elsewhere online compared to what Bard is telling you.
A further fresh introduction gives Bard users the ability to share a conversation via a public link, allowing others to continue that conversation with Google’s AI themselves, should they wish.
Analysis: The distant but distinct sound of alarm bells
This is indeed a major update for Bard, and there are some useful elements in here for sure. Better quality and accuracy, and the ability to check Bard’s responses, are obviously welcome features.
Some other stuff will set some alarm bells ringing for folks, particularly the more privacy-conscious out there. Do you really want Bard’s tendrils snaking into every corner of your Google Drive, Docs, and Gmail? Doesn’t that sound like the beginning of a scenario of a nightmarish overreach from the AI?
Well, Google is pretty careful here to clarify that your personal data absolutely isn’t being hoovered up to train Bard in any way. As the company puts it: “Your Google Workspace data won’t be used to train Bard’s public model and you can disable access to it at any time.”
So, the only use of the data will be to furnish you with convenient replies to queries, and that could be pretty handy. Know you’ve got a document somewhere on a certain topic, but can’t remember where it is in your Google account, or what it’s called? You should be able to prompt Bard to find it for you.
Don’t like the idea of Bard accessing your stuff in any way, shape, or form? Then you don’t have to use these abilities, they can be switched off (and the mentioned extensions don’t have to be enabled). Indeed, whatever assurances Google makes about Bard not snuffling around in your data for its own purposes, there will be folks immediately reaching for the ‘off’ switch in these cases, you can absolutely bank on it.