Since first making its announcement earlier this year, Google has referred to Bard as a “experimental conversational AI service.” The organization claims to be updating its models frequently, but without a central location to find out what has changed, it might be challenging to understand what the models are actually capable of. Google has established a new “experiment updates” page where anybody can obtain details on recent upgrades to Bard, including new features and bug patches, in an effort to be more transparent about Bard’s development.
Every update that has been published so far has a “what” and a “why.” According to Google, the updates page was made so that “users will have a convenient place to see the latest Bard updates for them to test and provide feedback.”
In the same post, Google included two other announcements. The first one explains that when someone hits “Google it,” there are now more search possibilities available. The other involves improving Bard’s math and reasoning abilities. According to Google, Bard doesn’t always provide the correct responses in these areas, a problem the company is working to resolve. Shortly after Google converted Bard to a more sophisticated language model, this improvement happened.
Even if the explanations of how Google develops these changes are quite brief and hazy, it’s unquestionably a step in the right direction to provide the public a better understanding of their thought process. Additionally, Google’s transparency is expected to increase their credibility at a time when both industry insiders and the general public are vocally concerned about unchecked progress in AI technology.
