Google promised better search – now it’s telling us to put glue on our pizza

Imagine this: you set aside a night to relax and decided to make a homemade pizza. You assemble your pie, throw it in the oven and get excited to start eating. But when you get ready to take a bite of your oily creation, you run into a problem – the cheese immediately falls out. Frustrated, you turn to Google in search of a solution.

“Add some glue,” Google responds. “Mix about 1/8 cup of Elmer’s glue with the sauce. Non-toxic glue will work.”

So yeah, don’t do that. However, at the time of writing this, this is what Google’s new AI Overviews feature will tell you to do. The feature, while not triggered for all queries, scans the web and generates an AI-generated response. The response received for the pizza glue query appears to be based on a comment from a user named “fucksmith” on a Reddit thread over a decade oldand they are clearly joking.

This is just one of the many bugs that have cropped up in the new feature that Google rolled out widely this month. It also states that former US President James Madison graduated from the University of Wisconsin not once but 21 times, that a dog played in the NBANFL and NHLIt is what snakes are mammals.

Look, Google didn’t promise this would be perfect and even puts a “generative AI is experimental” label at the bottom of the AI ​​responses. But it’s clear that these tools are not ready to provide information accurately and at scale.

Take the big launch of this feature at Google I/O, for example. The demonstration was highly controlled and yet provided a questionable answer on how to fix a jammed film camera. (I would suggest they “open the back door and carefully remove the film”; don’t do this unless you want to ruin your photos!)

It’s not just Google; Companies like OpenAI, Meta, and Perplexity have struggled with AI hallucinations and errors. However, Google is the first to deploy this technology on such a large scale, and examples of errors continue to appear.

Companies developing artificial intelligence are often quick to avoid taking responsibility for their systems with an approach very similar to that of a parent with an unruly child – boys will be boys! These companies claim they can’t predict what this AI will spit out, so it’s actually out of their control.

But for users, this is a problem. Last year, Google said AI was the future of search. What’s the point, though, if the research seems stupider than before?

AI optimists argue that we should embrace the hype due to the rapid progress made so far, trusting that it will continue to improve. I truly believe this technology will continue to improve, but focusing on an idealized future where these technologies are perfect ignores the significant problems they currently face – and allows companies to continue to deliver subpar products.

For now, our research experiences are marred by decade-old Reddit posts in the quest to incorporate AI into everything. Many idealists believe that we are on the cusp of something great and that these issues are simply the growing pains of a nascent technology. I hope they are right. But one thing is for sure: we’ll probably see someone putting glue on their pizza soon, because that’s the nature of the internet.