- Google is still dealing with the backlash caused by inaccurate answers from Gemini AI.
- Sergey Brin blamed a lack of testing amid reports that employees felt rushed to ship products.
- The incident illustrates the dangers of rushing products out as the AI arms race intensifies.
Google still hasn't been able to put out the Gemini fire.
The company has defended itself from intense backlash caused by its AI chatbot's image generation capabilities. The feature created historically inaccurate images of people of color, which the company later suspended and senior executives apologized for.
The product has placed Google at the center of what some circles describe as a “woke” culture war. Critics see the controversial model as evidence that Google's products are being unduly influenced by left-leaning employees.
Efforts by senior executives to apologize seem to have done little to calm many of its critics, many of whom wonder how the company could have shipped a product with such obvious problems. is doubtful.
Andres Gwirz, a lecturer at King's Business School, said Google's position in the intensifying AI arms race among big tech companies may have heightened the sense of tension within the company. told Business Insider.
“We're seeing a general trend of companies rushing to adopt and build AI solutions,” he said, adding that the pressure to bring products to market can lead to cutting corners.
“The engineers I talk to are well aware that this gold rush of integrating AI into existing software stacks comes with new legal and commercial risks, but they still decide when to ship new features. “We have limited authority to do so,” he said.
engineer in a hurry
One of the company's co-founders, Sergey Brin, addressed the complaints on Saturday, insisting that the model's leftward bias was not the company's intention and that executives are working to resolve the issue.
CEO Sundar Pichai also apologized for the incident and promised “structural reforms” in a memo to employees, but some have called for his resignation.
Brin said the strange images were likely due to a lack of thorough testing, and The Verge's Alex Heath said employees were being rushed to ship the product. He said he accused him of being there.
Heath cited the photo generation used in the Gemini app as an example of Google's haste. This utilizes an older text-to-image conversion model that was built into Gemini in order to roll out the feature faster.
The Verge also reported a lack of coordination between the research teams building the underlying models and the teams building them into products.
Google’s “Gemini era” is all about GPT-4
Pichai argued late last year that Google was entering the “Gemini era.”
This model is widely seen as the company's answer to OpenAI's GPT-4, a product that Microsoft has a lot of support for and is already included in many of its AI products.
Google is keen to keep up with its former rival, following Microsoft's lead in announcing similar products such as AI-powered search and workplace assistants.
Gemini's image generation failure isn't the first time Google has been publicly embarrassed about AI.
Last year, Google's Bard, an AI-powered chatbot widely seen as the company's response to ChatGPT, made an error during its first live demo. The mistake inadvertently highlighted the dangers of replacing search engines with chatbots and caused the stock price of Google's parent company Alphabet to plummet.
Similar failures occurred with Microsoft and OpenAI technologies. Days after Bard's failure, his AI-powered Bing from Microsoft began serving users “rambling responses,” including declarations of love and angry arguments.
But Google has worked hard to position itself as a responsible guardian of AI. In a Financial Times op-ed last year, Pichai said the “race to build AI responsibly” was more important than the need to ship user products.
For Google, the main issue in the Gemini controversy is brand awareness, Gwirtz said.
“Most people have traditionally perceived Google as 'neutral' and have never seriously asked themselves what it means to be a search engine provider,” he said.
Professor Sandra Wachter from the Oxford Internet Institute told Business Insider: “I think it's important to move away from the mantra of 'moving fast and solving problems' and focusing on 'moving at a responsible speed and solving problems.'” Technology is used to solve problems. and not just a way to beat others in the market. ”
Businesses do not exist in a vacuum, Wachter said, and their actions impact individuals and society.
“The world is not a sandbox or a large-scale field experiment, so we need to be careful about which products are brought to market and when,” she added.
A Google representative did not immediately respond to BI's request for comment.
On February 28, Axel Springer, the parent company of Business Insider, joined 31 other media groups in filing a $2.3 billion lawsuit against Google in Dutch court, alleging losses caused by the company's advertising practices. Ta.