Google Bard Launches to the Public: A New Beginning of Mistakes
Google announced the addition of ChatGPT-like features to its search engine called Bard and launched it for its ' trusted testers ' . Now, for the first time, Google has opened this feature to the general public. Of course , this feature is not yet available for public testing in our country as of this writing.
However, these features have been surrounded by controversy ever since. In the countries where its public testing has started, this chatbot made an overwhelming revelation during a conversation with users! The user asked the chatbot how it was trained . In response, the chatbot said that from the Gmail data of various users! When this came out, there was a lot of uproar. Even earlier, there was a controversy that Google company was reading our mails and showing advertisements according to its content. After that Google stopped that activity. Now the chatbot has revealed again that Google uses the data in our Gmail for its own benefit! However, Google has denied that Gmail data was used for bard training.
The funny thing is that the controversial conversation with Google Bard was done by a Microsoft researcher! When he asked the chatbot about its training, Bard replied that it gets its training from various sources on the internet, such as publicly available datasets , Wikipedia , GitHub , StackOverflow, etc. Bard then said that he got training from Google's internal sources like Google Search , Gmail and other products and also from some third party company data. The researcher shared screenshots of this conversation on social media. If it was indeed trained with Gmail data, it would be a huge breach of user data privacy. However, Google clarified that Bard is error-prone and is not trained on Gmail data!
Comments
Post a Comment