This Tuesday (21), Google launched to the general public in the US and UK its chatbot Bard, a competitor to the successful ChatGPT interface, with the goal of improving the quality of its answers through greater interaction with users.
The subsidiary of the Alphabet group at the beginning of February with some trepidation announced the creation of Bard, since in November ChatGPT appeared, a service from the startup OpenAI, created in cooperation with Microsoft.
Bard’s use was initially restricted to “trusted testers” but is now available to the general public. However, at the moment, the number of accesses is limited, there is a waiting list and only users in the US and UK are included.
It is platform independent of the Google search engine, with a text field where the user can type in questions.
“We learned a lot from testing Bard, and the next big step in improving it is getting feedback from more people,” Vice Presidents Sissy Hsiao and Ellie Collins wrote on a Google post.
“The more people use it, the better LLM will work,” they explained, referring to software capable of answering questions formulated in natural language.
Bard is built on top of LaMDA, a language model designed by Google for creating chatbots, the first version of which will be introduced in 2021.
Both Google executives acknowledge that LLM is “not perfect” and can “certainly provide inaccurate, misleading, or false information.”
Google says it has implemented “safeguards” to limit the possibility of a bot providing inaccurate or inappropriate responses, in particular limiting the interaction time between Bard and users.
Since the launch of ChatGPT, many have tried to push the chatbot to its limits and managed to generate ridiculous and even annoying responses.
In an interview with The New York Times, Hsiao and Collins acknowledged that Google had not yet established a business model and monetization strategy for Bard.
tu / juj / ag / llu / ic / mvv
