Google has made its generative AI platform, Bard, available to select users in the US and UK. The platform provides productivity tips, helps users understand complex topics and outlines blog posts, and displays three different drafts for each response. Conversations between users and the AI are capped to ensure interactions remain helpful and on topic. However, there are no limits to the number of daily chats. Users should use Google’s “Google it” feature to fact-check Bard’s responses as inaccuracies are a common issue with large language models. Google asked all employees to help fix Bard’s wrong answers in a company-wide effort.
Google Launches Bard AI Platform to Limited Users
Google has opened up its Bard AI platform to a select number of users in the US and the UK. Bard is Google’s generative AI product that helps users to boost their productivity, accelerate their ideas, and fuel their curiosity. With Bard, users can ask for tips to reach their goals, understand complex topics, or even spark their creativity by outlining a blog post. The company has been testing Bard for several months and is now expanding availability to more countries and languages. To access the product, users must sign in with a Google account.
According to a blog post by Sissie Hsiao, product vice president, and Eli Collins, research vice president, the next critical step in improving Bard is to get feedback from more people. Google is putting in place some guardrails to ensure that conversations between users and the AI do not spiral out of control. The company has capped the number of exchanges in a dialogue to keep interactions helpful and on topic. However, there is no limit to the number of daily chats.
Google expects the testing phase to help it understand how often people engage with Bard. The AI will display three different drafts for each response. The launch comes after months of rapid testing following the introduction of OpenAI’s ChatGPT technology in November, which has since been integrated into a limited version of Microsoft’s search engine Bing.
Google’s LaMDA technology is also integrated into Bard chatbot and on search pages, and the company announced the products in January 2023. Last week, Google said it would open its AI products to a limited number of users of Workspace, which includes Gmail and Google’s productivity tools.
Expect Mistakes
Google is warning users that Bard may make mistakes, and they should use the “Google it” feature to fact-check Bard’s responses if they’re in doubt. The company is currently working on resolving inaccuracies, a significant issue with large language models (LLMs). Google executives asked all employees to help fix Bard’s wrong answers in a company-wide effort.
“While LLMs are an exciting technology, they’re not without their faults,” Google’s blog post said. “For instance, because they learn from a wide range of information that reflects real-world biases and stereotypes, those sometimes show up in their outputs. And they can provide inaccurate, misleading, or false information while presenting it confidently.”
Leave a Reply