Google seems to be enthusiastic about OpenAI’s ChatGPT. Artificial intelligence-powered chatbots have taken the tech world by storm in recent months because they can provide users with the information they’re looking for in an easy-to-understand format.Google sees ChatGPT as a threat to its search business. , has changed plans accordingly over the past few weeks. new york times.
The report claims that CEO Sundar Pichai has declared “Code Red” to accelerate AI development. Google is reportedly preparing to show off at least 20 AI-powered products and chatbots for its search engine this year, at least some of them at his I/O conference in May. I am planning to debut.
According to the slide material viewed by Times, some of the AI projects Google is working on include an image generation tool, an upgraded version of AI Test Kitchen (the app used to test prototypes), a TikTok-style green screen mode for YouTube, and other recap videos There are tools that can generate clip. Also in the pipeline is a feature titled Shopping Try-on (presumably similar to one Amazon is developing), a wallpaper creator for Pixel phones, and an easy way for developers to create his Android apps. We have AI-powered tools to help.
Pichai reportedly invited Google founders Larry Page and Sergey Brin last month to meet with current leaders to review and voice their AI plans. The duo don’t have much day-to-day involvement with the company since 2019, as they’ve focused on other projects.
Google claims it has been trying to speed up its product approval process, including checks to ensure its AI-driven technology is fair and ethical. Search chatbot demo priorities include safety, accuracy, and blocking misinformation It seems that. However, for other products and tools Google is working on, it “has a lower bar and seeks to contain rather than prevent issues related to hate, harm, danger and misinformation,” it said. Times report.
Recently, Google has been careful when announcing new products. The slide reportedly mentions “copyright, privacy and antitrust” as the main risks of AI technology. It is said to have pointed out the need for solutions to eliminate copyrighted material and prevent personally identifiable information from being shared.
Over the past few years, there has been backlash over Google’s handling of AI ethics. Timnit Gebru and Margaret Mitchell, two leading AI ethics researchers, said Google fired them for censoring research that criticized her AI language learning model. criticized. This includes concerns that AI language learning models encode biases found in training data. The result could be “a model that encodes stereotypical and derogatory associations along gender, race, ethnicity, and disability status,” the researchers wrote in their paper. The training dataset can also contain incorrect information. After Gebru and Mitchell left, her two other prominent ethics researchers left Google early last year.
It’s not hard to see why Google is said to be panicking on ChatGPT. For one, a report earlier this month suggested that Microsoft (an investor in OpenAI) plans to incorporate some of the technology that powers ChatGPT into his Bing. The company announced this week that it will soon integrate ChatGPT into the Azure OpenAI Service.
The latest report on Google’s response to ChatGPT came out shortly after Google announced it would lay off 12,000 people. “Thanks to the strength of our mission, the value of our products and services, and our early investment in AI, we are confident in the great opportunities ahead of us,” Pichai wrote in a note to staff. wrote in. “To fully capture it, you have to make some tough choices.”
The CEO said the company is preparing to “unveil an all-new experience for users, developers, and businesses. And we are ready to tackle it responsibly,” he added.
All products recommended by Engadget are selected by an editorial team independent of the parent company. Some stories contain affiliate links. When you purchase something through one of these links, we may earn an affiliate commission. All prices are correct at the time of publication.
