Google AI Grammar Trainer

Mountain view In the technological race for artificial intelligence, Google chief Sundar Pichai has spoken out in favor of legal guidelines for new technologies. In particular, he mentioned the possibility of creating so-called fake videos in which real people can be depicted in a deceptively real way.

“There should be consequences for producing deepfakes that harm society,” Pichai said in an interview with CBS’s 60 Minutes.

Experts have long warned of the danger of phishing with the help of so-called generative artificial intelligence (AI), which can produce texts, images or videos. These fears have recently been reinforced, for example, by an artificial image of the Pope and a series of images created by Amnesty International of the alleged arrest of former US President Donald Trump.

The software creates its own content based on the massive amounts of information it has processed for training. Using text, for example, programs like ChatGPT guess word for word how a sentence should go. Currently, one consequence of this principle is that they can sometimes output information that is completely incorrect – without the user realizing it.

A call to “social norms”

“Anyone who’s worked with AI for a while understands that it’s so different and so pervasive that we need societal norms to think about how we adapt,” Pichai said. One does not wish to bring any technology into the public domain for which society is not prepared.

The current AI race among the major technical players started with the launch of ChatGPT by OpenAI at the end of last year. Google has been working on various programs with artificial intelligence for years, but has repeatedly hesitated to make them available to the public, citing social responsibility. Meanwhile, an agreement between OpenAI and Google’s rival Microsoft has put the Internet giant under pressure.