On Tuesday, Google released its ChatGPT rival AI tool Bard to the public in an effort to improve the quality of its responses through greater user engagement.
The tech giant, owned by the Alphabet group, announced the creation of Bard in early February in response to the November launch of ChatGPT, developed by startup OpenAI in collaboration with Microsoft.
Being able to create “emails”, dissertations or lines of code on a simple request, ChatGPT is sparking a huge enthusiasm for generative artificial intelligence.
Bard’s use was initially limited to “trusted testers” before it was opened to the general public on Tuesday.
However, the number of connections was limited and a waiting list was created to manage demand.
Access is currently only available from the US and UK.
“When people start using Bard and testing its capabilities, they will surprise us,” said Google CEO Sundar Pichai in a message to his team prepared by Agence France-Presse (AFP).
“Something will go wrong. But user feedback is essential to improve the product and the underlying technology,” he added.
The leader of the California group has come under fire from within for launching Bard too quickly to catch up with Microsoft.
The interface consists of a “site” other than the Google search engine, with a space where the user can enter a question.
When asked by AFP about what makes it different from ChatGPT, Bard said that, unlike its competitor, it “is able to access real information through the Google search engine.”
“Chat-bot” [robô de conversação] he also emphasized that “it is still under development and ChatGPT is already available to the general public.”
“This means that I am constantly learning and improving, and ChatGPT will definitely remain unchanged,” he emphasized.
Already Google VPs, Sissy Hsiao and Eli Collins, stressed in a post on the group’s “website” that they learned a lot to test Bard.
“The next big step in improving it is to get ‘feedback’ from more people,” they stressed.
“The more people use it, the better the large language models (LLM, a program that can generate answers to questions formulated in everyday language) will be better at predicting answers that can be useful,” the two managers explained.
By receiving written data and conversations, the LLM algorithm can more accurately determine the appropriate answer to a question.
Bard uses LaMDA, a language model developed by Google to build “chatbots”, the first version of which was launched by a group in Mountain View, California in 2021.
Two Google executives acknowledge that LLMs are “not perfect” and may “reliably provide inaccurate, misleading, or false information.”
Google indicates that it has implemented “protections” to limit the possibility of inaccurate or inappropriate responses, in particular, limiting the duration of exchanges in the dialogue between the Bard and the user.
Since the launch of ChatGPT, several users have tried to push the “chatbot” to the limit and have generated absurd, even disturbing responses.
In an interview with The New York Times, Sissy Xiao and Eli Collins said that Google has yet to define a business model and strategy to profit from Bard.