As it approaches its first birthday, Google’s conversational AI tool Bard is getting another upgrade.
Early reviews found that Bard lags behind competing chatbots in part because its responses are less human-like. Since then, Google has integrated its AI model Gemini. This allowed Bard to expand training from text to video, audio, and photos. Currently, generative AI remains in the spotlight and is poised to make further advances in practicality.
in blog post On Thursday, Bard’s head of product Jack Krawczyk said the tool now allows users to generate images for free.
When someone types in a prompt like “Create an image of a hot air balloon flying over a mountain at sunset,” Bard generates what Google describes as “a wide range of custom visuals to help bring your ideas to life.” .
However, it takes a while (about 13 seconds to be exact).
Also, while most of the queries we tried produced relevant images and responses, they were not 100% accurate.
For example, when I asked Bard to create an image for a news article about the latest update to a tool, it refused to generate the image I requested. Then I simply asked Google to create an image about Bard, and it created this blonde cyborg.
Last year, the market was flooded with chatbots like OpenAI’s ChatGPT, Microsoft’s Bing AI, Anthropic’s Claude, and yes, Google’s Bard, as big tech companies looked to stake their claim on the next wave of search. Masu. These chatbots access huge datasets and use large language models to provide responses to consumer queries with text, even images, and even video. This is rapidly evolving and is already approaching human conversation. However, while bots may provide answers with confidence, they are not necessarily accurate. It is also still vulnerable to exploitation.
Google’s post noted that Bard includes a distinction between visuals created with Bard and original human artwork, and that watermarks are embedded in the pixels of the images produced. To test this, we asked them to create an image of Botticelli’s The Birth of Venus. I provided a replica, but it was sloppier. Those faces! That hand! However, you have the option to report legal issues and give each image a thumbs up or thumbs down.
In the wake of the Taylor Swift deepfake, Google is working to limit “violent, offensive, or sexually explicit content” and will apply filters to avoid generating images of named people. He said he is doing so. In fact, he refused to create images depicting Super Bowl quarterbacks Patrick Mahomes and Brock Purdy having a picnic or Beyoncé in a bank.
“We continue to invest in new technologies to improve the security and privacy protection of our models,” Krawczyk wrote.
When I asked Bard to generate an image of Lisa Lacy at work, he said he didn’t have enough information about the person to be useful. But I was able to create an image for a more common question from a journalist at work: there are two sandwiches on the desk instead of her one.
He declined to create an image of a man tossing coins from Hoover Dam because “throwing objects into Hoover Dam is prohibited.” (He offered in return for him to create images of landscapes and historical depictions.)
And Mr. Byrd happily created images of historical moments such as the signing of the Declaration of Independence.
In addition to adding image generation tools, Google is expanding the availability of Gemini Pro in Bard from English to over 40 languages. This includes a double-check feature that allows users to fact-check his Bard’s responses using her web content.
This will expand the tool to more than 230 countries and territories, according to the post.
Google first added Gemini Pro to Bard in December 2023, offering “more advanced understanding, reasoning, summarizing, and coding capabilities.”
Editor’s note: CNET uses an AI engine to create some stories. For more information, see this mailbox.