We hate this timeline.
Cursed Ping Pong
Here's a nightmare scenario for our digital futures: the internet becomes so laden with AI-generated misinformation that its nearly unusable, in part because chatbots and other AI systems keep volleying hallucinated, mangled, or otherwise false or misconstrued information back at one another until our entire information landscape is shattered into tiny, blurry pieces and nothing is real anymore.
We're not there quite yet. But we did just get a glimpse of what that might look like.
Yesterday, The Verge had the unfortunate duty of reporting that, when asking Microsoft's chatty Bing AI a simple query, it responded by citing misinformation generated by Google's freshly-released Bard chatbot. And mind you, this all happened just one day after Bard's release. And away we go!
The digital pathway that Bing took to get to its wrong answer was dizzying — not to mention a perfect illustration of how poor the media literacy of these bots can be.
As explained by the Verge, when asked yesterday if Google's Bard had already been shut down, Bing told one user that yes, it had, "citing as evidence a news article that discusses a tweet in which a user asked Bard when it would be shut down and Bard said it already had, itself citing a comment on Hacker News in which someone joked about this happening, and someone else used ChatGPT to write fake news coverage about the event."
Oh, and that first news article even called Bard's original mishap an "obvious mistake." Even so, Bing repeated its claims.
It's true that humans aren't always top-notch when it comes to media literacy. We make mistakes, and to the credit of Microsoft, OpenAI, and Google, all of these companies have been honest about the fact that their products will make mistakes, too.
But that said, humans generally aren't built by Big Tech and marketed as information-gathering and synthesizing search assistants. No matter how many disclaimers these companies provide, a lot of people are sure to take what AI chatbots say at face value.
In any case, be wary. Information, after all, is still the internet's greatest — or at the very least, its most necessary — currency. And for all their impressive qualities, it's important to remember that providing accurate info is still the thing that these AI systems are worst at.
READ MORE: Google and Microsoft’s chatbots are already citing one another in a misinformation shitshow [The Verge]