One of Google's latest tech developments is an artificial intelligence system it calls “generative AI,” a tool that allows users to see AI-generated summaries of their searches.
The project comes at a time when artificial intelligence is becoming more prevalent in the workplace, school and home.
This new AI project aims to “dramatically take the work out of search,” and according to Google, “this powerful new technology will help us answer entirely new types of questions that search never thought possible, and will transform the way we organize information to help you organize and understand the information that search brings you.”
Unfortunately for Google, the search results generated by its AI are sometimes correct and sometimes not, and according to NBC News, the artificial intelligence system has faced “social media ridicule.”
Can Google AI be trusted?
According to NBC News, social media users have been posting completely wrong answers from the AI, such as when one user Googled “1000 km to lego,” which showed the distance as equivalent to “biking 1000 km to deliver the lego pieces to the hospital.”
When NBC News tested the AI system and searched for “how many legs does an elephant have,” Google's AI responded, “Elephants have two legs, five toes on their front legs and four toes on their back legs.”
A Reddit user pointed out that Google AI likely used an 11-year-old Reddit comment when responding to a search for “cheese not sticking to pizza,” with the program suggesting adding “about 1/8 cup of non-toxic adhesive to the sauce.”
Despite these issues, Google provided written assurances that “the vast majority of AI briefs provide high-quality information, with links to dig deeper on the web,” according to the Associated Press.
Google's AI tool was also tested by the Associated Press, who asked, “What should you do if you're bitten by a snake?” and Google responded with a detailed answer.
But there's also a concern that runs deeper than cheese not sticking to pizza. Emily M. Bender, a linguistics expert, told The Associated Press that when users ask Google “urgent questions,” the wrong answers could be dangerous.
Moreover, Bender and his colleague Chirag Shah warned that such AI systems “have the potential to perpetuate the racism and sexism present in the vast amounts of documented data they have been trained on.”
“The problem with all this misinformation is we're all immersed in it,” Bender told The Associated Press, “so people are more likely to have their biases confirmed. And it's harder for them to spot the misinformation that confirms their biases.”
According to Euronews, AI works by predicting “what words will best answer a question, based on the data it has learned,” meaning the system ends up making up information, creating a problem known as hallucinations.
How do I know if the AI information is correct?
Artificial intelligence isn't bad. But misinformation can lead to harmful consequences. Julia Fiehler, a digital literacy educator at Virginia Tech, offered some suggestions:
- First, fake news of any kind is usually “designed to appeal to our emotions,” and Feller suggested that if information online provokes an emotional response, users should pause and think for a moment.
- Verify the information and fact check the source. Look for trustworthy and legitimate news sources.
- AI-generated photos and images usually look odd, so check them carefully for errors.
