Google’s conversational AI chatbot Bard failed to impress when it provided the wrong answer in its first public demo. Users were similarly underwhelmed when Google opened up access to Bard late last month, as the chatbot failed to perform as well as OpenAI’s ChatGPT. Google is still playing catch-up, but Bard’s first big update should give the AI a boost.
Google rolls out Bard’s first experiment update
On Monday, Google launched the inaugural experiment update for Bard. The update sees the arrival of a new “Experiment updates” page, where Google will share all of the latest features, improvements, and bug fixes for the Bard experiment.
If you want to keep track of all the changes Google makes to Bard in the coming months, you might want to bookmark that page in your browser.
Google also added more suggested Search topics when users click the “Google it” button that appears below Bard’s responses. This should make it easier to find stories about the subjects you’re interested in if you want to know more than what Bard offers.
Finally, Google updated Bard with better capabilities for math and logic. As Google explained, Bard frequently struggles to provide high-quality responses to questions that test its math and logic skills. Bard’s performance should improve following the update. This upgrade can likely be attributed to Google incorporating PaLM (Pathways Language Model) into Bard, which Google product director Jack Krawczyk discussed on Twitter two weeks ago:
It is unclear how often Google will update Bard, but we do know that Google has dedicated a significant number of resources toward this experiment. The next step will be integrating Bard directly into Google Search, but it might need another update or two first.
If you haven’t already, you can join the waitlist for Bard at bard.google.com.
