Tech giant Google has released limited access to Bard, its rival ChatGPT chatbot, marking an important step in its bid to regain what many see as lost ground in a new race to implement artificial intelligence (AI).

Although Google is emphasizing that Bard is an experiment and not a finished product, the platform is already available to select users in the US and UK, with the ability to join a waiting list at bard.google.com .

Bard, a plugin for search

Like Microsoft’s ChatGPT and Bing chatbot, Bard offers users a blank text box and an invitation to ask questions on any topic they like. However, given the well-documented tendency of these bots to fabricate information, Google is emphasizing that Bard is not a replacement for its search engine, but rather a “search plugin,” a bot that users can talk to. about life, generate writing drafts or simply exchange ideas.

Here is an example question

And this is the result:

In the end, it offers the possibility of doing a search in the good old Google:

An experiment to boost productivity and creativity

In a blog post written by two of the project’s leaders, Sissie Hsiao and Eli Collins, they describe Bard in cautious terms as “an early experiment… intended to help people boost their productivity, accelerate their ideas, and feed their curiosity.” ». They also characterize Bard as a product that allows users to “collaborate with generative AI” (emphasis ours), language that seems intended to diffuse Google’s responsibility for future bot outbursts.

Here you can see how he gives different advice on the same question:

The company views Bard as a automatic generation tool for ideas, emails and poems, rather than a tool to find information. Although, the company has been clear that if you want to search for information, you should use the “Google it” button that appears at the bottom of most of Bard’s responses.

Limitations and challenges

However, like any other technology, Bard has its limitations. Although the chatbot is connected to Google search results, it cannot always provide accurate information. In a demo for The Verge, Bard was able to quickly answer several general questions, such as suggesting ways to encourage a child to bowl and recommending a list of popular heist movies. But although it generated three responses to each user query, the variation in its content is minimal.

Bard also lacks Bing’s clarity around tagged footnotes, which Google says only appear when directly citing a source like a news article. Also, Bard seems to be more limited in his responses, which can be seen as both a blessing and a curse. On the one hand, Bing’s tendency to go off script secured a spot on the front page of the New York Times, but on the other hand, it can lead to inaccurate responses and misinformation.

Nonetheless, Bard is an interesting tool for generating creative ideas and irrelevant questions.such as movie or restaurant recommendations. Also, it’s important to note that Bard is an experiment, so Google is working to improve it.

As for your future, it may be integrated into Google search results, since it is consistent with the company’s vision that users seek adventures more than answers. Bard could be a useful tool to offer ideas and generate inspiration for users.

Challenges in AI implementation

Bard is based on LaMDA, Google’s AI language model, which is much more capable than this limited interface suggests. However, this also raises issues for Google as to how much of that potential it should expose to the public, and in what form. Bard needs to expand his repertoire if he is to be heard, but it also needs to be carefully crafted so as not to generate harmful or dangerous responses. The risk of users encountering inaccurate or offensive information is real, and while Google has added warnings and disclaimers to Bard’s interface, this may not be enough to prevent problems.

Google’s launch of Bard is yet another sign of the tech industry’s growing interest in AI and its potential to change the way we interact with technology. However, it also demonstrates the challenges and risks associated with implementing AI in everyday life. It is important that companies are transparent and cautious in the way they present and market these technologies, and that users are aware of their limitations and risks.