Google today announced early access to Bard, an AI-powered chatbot designed to boost creativity and productivity. Users in the U.S. and U.K. can sign up to try Bard, which is built on Google’s LaMDA language model and provides open-domain responses to user questions.
Bard is designed to increase productivity, catalyze thinking and inspire imagination. Users could ask the AI to provide recommendations for accomplishing a reading goal, explain complex concepts in lay terms or generate an outline for a blog post. Early testing has yielded insights to strengthen Bard, and Google says gathering diverse user feedback is essential to continued progress. By interacting with and evaluating Bard, the public can help optimize the system and shape its development.
Bard will often provide multiple response options to choose from. The system continually learns from interactions, but Google acknowledges the challenges of AI language models, including biases and inaccuracies. Users can click “Google It” buttons to double check responses or search the web for more information. The company says Bard is designed to complement, not replace, web search.
“Bard is an early experiment that lets you collaborate with generative AI,” according to Google. The technology is grounded in Google’s understanding of quality information, with responses generated word-by-word based on probabilities. Google will update Bard over time with newer AI models and plans to expand availability to more countries and languages. Those interested in trying Bard can sign up at bard.google.com.