The fall of ‘Google-fu’

What’s the best way to overcome a vexing tech problem? For years, IT professionals turned to Google (or their search engine of choice) to find a solution. Reddit, Stack Overflow, and other forums are filled with the answers to some of the worst IT crises you can imagine.

Those IT pros proficient at “Google-fu,” a term for martial arts-like googling skills, know how to use specific keyword combinations to yield the results they want, fast.

But generative AI might be rendering Google Fu obsolete—and creating new problems besides. As more IT pros turn to ChatGPT, Gemini, or other LLMs for answers that may or may not be correct, “AI slop” dominates the top of some search engine results, offering the “worst of both worlds,” according to Tom Bachant, the founder and CEO of Unthread, an AI-powered support platform that helps with resolving help tickets within Slack.

“In some cases, you don’t get as comprehensive of a response as you’d get if you ask Claude or ChatGPT directly,” Bachant said. “You also get a bunch of links and ads that maybe aren’t relevant to your question. So, if you’re looking for a clear answer that’s written in plain English, you’re gonna get a better version of that through an LLM than you will out of the given Google AI search results.”

The fall of Google-fu. Brian Alvey, the CTO of WordPress VIP, told IT Brew that AI slop in search results will drive users to rely more on recognizable brands, forcing other websites to compete for visibility.

But even those recognizable brands might end up buried. “What you’re seeing today is, on a page-by-page basis, it’s very hard for Google to see…a good, reputable recipe result or a good, reputable Stack Overflow code problem, DevOps ops solution result,” Alvey said. “But Google will figure this out in the long term, which could be weeks from now.”

Is it that bad? Search results can still lead a user to the source of good information, but this often involves a very human judgment call. Jody Bailey, the chief product and technology officer at Stack Overflow told IT Brew that it can become difficult to know what to rely on.

“As an IT professional, you tend to weigh the risk and say, ‘Okay, well this feels close enough that I can go with it,’ or ‘I need more information, I just need to find it someplace else,’” Bailey said.

Isn’t there any way you trust me? Bailey said that some of his colleagues have put search engines in the rearview. Instead, they’re asking chatbots for help.

“I think what’s really important…is providing that attribution, a way to link back to the data so that people can actually discern for themselves,” Bailey said.

He added that if an IT pro used an LLM or the AI slop at the top of search engine results to troubleshoot or brainstorm, he’d accept it because “people get ideas and solutions from all sorts of different places.”

“If he was just blindly trusting something that he read someplace on the internet, I’d have doubt,” Bailey said. “I still believe human judgement is essential.”