Google had removed a feature that allowed AI chatbots to generate portraits of people, but after significant backlash, the search giant has now reintroduced the feature.
Google announced on Wednesday that it is rolling out an updated image generation tool to English-language users of its AI chatbot, Gemini Advance, which will also feature Google's latest image generation model, Imagen 3.
In February, Google disabled users' ability to generate images of people after acknowledging that its tools “over-correct” to portray diverse people in unrealistic or absurd scenarios, highlighting the company's desperate efforts to jump on the AI bandwagon and the risks of releasing new technology without rigorous vetting.
Images portraying black women as America's founding fathers and black and Asian people as Nazi-era German soldiers drew heavy criticism on social media.
Some users argued that the results the tool produced were evidence of anti-white bias.
The viral image drew criticism from prominent figures, including Google Inc. owner Elon Musk, who has criticized AI tools for being biased toward liberals. Google CEO Sundar Pichai told staff at the time that some of the missteps were unacceptable and “we were wrong.”
“With Imagen 3, we've made significant contributions to improving the user experience when generating images of people,” Dave Citron, a senior director at Google, said in a blog post on Wednesday.
The reintroduction is the latest development in Google's sometimes uneven AI rollout.
Last year, Google made a gaffe when it released its first AI chatbot, Bard, which spit out inaccurate responses in a promotional video.
Taken together, these incidents highlighted a sloppy launch and undermined Google's credibility at a time when tech giants were jockeying for position amid the AI boom.
Google has announced it is upgrading its creative image generation capabilities, introducing its latest image generation model, Imagen 3, to Gemini Apps. (AP Photo/Juliana Yamada) (Associated Press)
To prevent abuse, Citron said the tool does not support generating graphic scenes of identifiable individuals, minors or excessive violence.
“Of course, like any generative AI tool, not every image Gemini creates is perfect, but we're listening to feedback from our early users and will continue to improve it. We aim to gradually roll this out and make it available to more users and languages soon.”
Google's relaunch comes at a challenging time in the AI world.
In the months since the search giant first released and then pulled the tool, investors have been wavering in their support for AI investments, with recent market volatility resurfacing questions about the valuation of tech companies that have poured huge resources into ambitious AI builds.
The story continues
The competitive environment has also changed.
While Google was previously thought of as playing catch-up with Microsoft (MSFT), Apple has since joined the big tech AI camp. And while the iPhone maker's rivals have been dealing with boardroom squabbles and public backlash over their AI developments, Wall Street has praised Apple's efforts, highlighting how it has successfully marketed its AI tools in a way that other tech giants have struggled to do.
Google's reintroduction of the image generation feature comes just before a big earnings announcement from Nvidia (NVDA). Shares of the AI chip designer have risen about 160% so far this year, and AI bulls see the blowout earnings report as a possible next catalyst in what has been a tough market for chip and AI companies.
Hamza Shaaban is a reporter for Yahoo Finance covering markets and economics. Click the X to follow Hamza. translation:.
For the latest technology news impacting the stock market, click here
Read the latest financial and business news from Yahoo Finance
