Google Bard, the eagerly anticipated generative AI chatbot from the search engine giant just inched closer to full-public release. It’s now an open beta.
That’s the good news; the bad news is that there’s already a waiting list for Google Bard and, while you can sign up now with a personal Google account (US and UK, only), there’s no telling when you’ll gain access to the conversational AI system.
Like OpenAI’s ChatGPT, Google Bard is a generative AI system – that’s to say, it’s been trained, using vast amounts of human-created online content, to accurately predict the right words in the context of a response.
Google unveiled it weeks ago in a less-than-inspiring demo, with some observers suggesting Google had been caught flat-footed after Microsoft’s blockbuster AI-enabled Bing (with the ChatGPT-powered co-pilot) launch in February.
Now, though, Google Bard appears a little less half-baked. Google announced the still somewhat limited availability in a blog post – which, naturally, Bard helped write – on Tuesday, explaining:
“You can use Bard to boost your productivity, accelerate your ideas and fuel your curiosity. You might ask Bard to give you tips to reach your goal of reading more books this year, explain quantum physics in simple terms or spark your creativity by outlining a blog post.”
Google Bard is based on Google’s LaMDA large language model (LLM); yes, the one that was accused of achieving sentience. It’s not, however, the best or smartest version of LaMDA. Instead, Google is using what it calls a lightweight version, and will update it with a smarter one later. Think of this LaMDA like OpenAI’s GPT-3, and the next version as a sort of GPT-4.
In images provided by Google, Bard appears to work similarly to ChatGPT. There’s a text-prompt field where you can enter the query of your choice, and at the top is a welcome message from Bard, where it proclaims, “I’m Bard, your creative and helpful collaborator. I have limitations and won’t always get it right, but your feedback will help me improve.”
Google advises users that Bard is not without fault, and cautions that an LLM can “provide inaccurate, misleading or false information while presenting it confidently.”
In other words, if and when you do gain access you may want to take Google Bard’s advice and information with a grain of salt – which is probably good advice for using any chatbots.
Where Google Bard seems to differ from ChatGPT is that it’s built from the getgo for collaboration, offering not just one response, but multiple drafts you can choose from. This creates a sort of LLM branching effect, where you’re ultimately choosing your own AI content-building journey.