We had already talked about it google bard, but now the new chat service signed by Mountain View has made its official debut in Great Britain and the United States. In this in-depth analysis, we will see what the new Google chatbot is and how to use it, which is the main competitor of ChatGPT.
Topics of this article:
Google Bard: what it is
When it comes to online search or artificial intelligence (AI), Google is usually the point of reference. However, today the tech giant has distinguished itself by launching Bard, a chatbot designed to compete with the very popular ChatGPT.
Bard, just like ChatGPT, is also capable of answer user questions need discuss a wide range of topics, showing at times an understanding comparable to the human one. Google demonstrated several of the chatbot's responses to Wired US, including recommendations for bowling activities for a child and a list of twenty recommended books for this year.
A trait in common between Bard and ChatGpt is the propensity, on some occasions, to create imaginary situations and to adopt bizarre behaviors. An example cited by Google concerns the chatbot's inability to correctly identify a plant species, which is suggested for growing at home. “Bard represents an early experiment, so he can't be perfect and will sometimes make mistakes“explained Eli Collins, vice president of Google's search division, which is in charge of the Bard project.
According to reports from Google, Bard has been made available to a small group of testers, but from today anyone in the US and UK will be able to request access to the system.
Google Bard: how to use it
The chatbot will be accessible through a dedicated web page, which will be distinct from Google's standard search interface. Bard will provide three answer options for each question, with the aim of making it clear to users that the system can generate answers in seconds, even if it can sometimes make mistakes.
In addition to each response from Bard, Google will also suggest a related traditional search query. Additionally, users will be able to provide feedback on responses through a thumbs-up or thumbs-down rating system and enter a more detailed comment to help the company improve the chatbot.
Google reported that early adopters of Bard found it useful for generating ideas or text. Collins admitted that some users were able to make the chatbot act improperly, though he didn't specify what restrictions Google placed on the system.
Meanwhile, new chatbots are springing up like mushrooms…
In 2020, Google had presented for the first time a chatbot developed with technologies similar to those used for ChatGpt. However, the company preferred to proceed with caution, especially after one of its engineers caused a significant media outcry by claiming that the language model he was working on had become "sentient". The acceleration in the Bard project demonstrates that the excitement generated by ChatGpt has prompted Google to take greater risks.
In the month of February, poco after investing ten billion dollars in the company that developed ChatGpt, OpenAi, Microsoft has launched a conversational interface for its search engine Bing, based on the same technology. Also, the Chinese company Baidu has announced its chatbot, Ernie Bot, earlier this month.
The race to develop and commercialize technology appears to have accelerated. Last week, OpenAi announced an improved version of the language model behind ChatGpt, called Gpt-4. Instead, Google has announced that it is making available a powerful language model called PaLM and to add text-generation features to its enterprise software, Google Worksplace.
