Google Introduces Bard: The New AI Chatbot to Watch

Google has announced the launch of Bard, its new AI-powered chatbot, which is set to compete with similar products offered by Microsoft and OpenAI. Bard will initially be available to a select number of users in the US and the UK, but those who are interested can join a waitlist at bard.google.com. Google has stated that it will take a slow and steady approach to the roll-out and has not yet provided a date for full public access.

Bard will operate in a similar way to its competitors, offering users a text box in which to ask any question they like. However, Google has emphasized that Bard is not intended to replace its search engine, but rather to complement it. The chatbot can be used to generate writing drafts, bounce ideas off of, or simply chat about life. In a blog post, the project leads, Sissie Hsiao and Eli Collins, describe Bard as “an early experiment” intended to help people boost their productivity, accelerate their ideas, and fuel their curiosity. They also characterized Bard as a product that lets users “collaborate with generative AI,” a phrase that seems intended to diffuse Google’s responsibility for future outbursts.

During a demo for The Verge, Bard was able to answer a number of general queries quickly and fluidly, offering advice on how to encourage a child to take up bowling and recommending a list of popular heist movies. Bard generates three responses to each user query, though the variation in their content is minimal, and underneath each reply is a prominent “Google It” button that redirects users to a related Google search. As with ChatGPT and Bing, there’s also a prominent disclaimer underneath the main text box warning users that “Bard may display inaccurate or offensive information that doesn’t represent Google’s views.”

While Bard is connected to Google’s search results, extracting factual information from it can be hit-and-miss. For example, it couldn’t fully answer a query on who gave the day’s White House press briefing or a tricky question about the maximum load capacity of a specific washing machine. Although repeating the query did eventually retrieve the correct information, users would be unable to know which answer was correct without checking an authoritative source like the machine’s manual.

Bard is faster than its competitors, though this may be because it currently has fewer users. It seems to have potentially broad capabilities, including generating lines of code, but it also lacks Bing’s clearly labeled footnotes. For Google, this could be both a blessing and a curse. While Microsoft’s Bing received negative attention when its chatbot was seen alternately insulting, gaslighting, and flirting with users, it also endeared the bot to many. Bard doesn’t seem to have any of that chaotic energy.

As more users gain access to Bard, this collective stress test will reveal the system’s capabilities and liabilities. One attack that remains untested is jailbreaking, which involves inputting queries that override a bot’s safeguards and allow it to generate harmful or dangerous responses. Bard is based on Google’s AI language model LaMDA, which is much more capable than this constrained interface implies. Google must decide how much of this potential to expose to the public and in what form. Given initial impressions, Bard needs to expand its repertoire if its voice is going to be heard.

Source: Techmeme