Google’s AI Search Results See Some Fine-Tuning

Has Google been telling you to put glue on your pizza, run with scissors, or eat “at least one small rock a day”? If you have been a victim of some interesting AI advice you might be happy to hear that you should be seeing a lot less now.

A recent study by BrightEdge has reported the number of AI-generated search engine results has dropped in recent weeks, which is a plus for many. So why is Google saying they don’t agree?

The Numbers

The study by BrightEdge showed AI-generated results dropping from 11% to 7% between the dates of June 1st and June 30th respectively. On top of that, they claim that the AI results that are provided don’t cite Reddit or Quora as often (who were the major culprits behind the misinformation provided). Besides taking up space at the top of your search, there are other reasons these numbers are being looked at. Content creators and those who work in SEO (Search Engine Optimization) have claimed this new AI has made it harder to gain organic traffic.

Ultimately, this study hints at the fact that Google’s AI is simply not up to the task and is being pulled back. Google’s Take Google (through their spokesperson Ashley Thompson) has not only questioned the study’s methodology but also strongly contested the results. They claim that the study doesn’t reflect what they are seeing on their end. They also claim that BrightEdge was mixing people who have opted in to “AI Overviews and More” with those who haven’t.

Opting in will allow you to see more AI-generated results while not opting in will obviously not push it as much. Mixing users from both categories could make it appear as if AI-generated results are dropping. BrightEdge later clarified that they did not do that and only used users who have opted in to “Ai Overviews and More”.

However, Google is doubling down. They continue to say their results are “helpful and provide value beyond existing features on the results page.” And while they admit to some fine-tuning, they stand strong that this has not translated into less AI-generated results in total. So what kind of fine-tuning are they doing?

One, limiting user-generated content which should lower absurd answers thanks to AI being unable to detect sarcasm. Two, improving the detection of “nonsensical queries” or basically, trick questions asked with the intent of making AI give a bad answer.

What Does This Mean For You?

While you might be able to quickly find the error when asked to put glue on your pizza, these inaccurate AI results have some serious consequences for those who make a living with the help of Google.

Sometimes fake facts aren’t always that easy to see, something journalists should keep in mind. Fact-check your sources, always, even if they seem to come from an authoritative source such as a search engine. Keeping on top of changes in Google’s AI and usage rates, in general, can also help marketers and digital professionals alike. This might look like more frequent SEO research or article rewrites or simply more research into what users want that AI can’t provide.

You’ve heard this before, but the digital landscape is always changing. Now more than ever. Love it or hate it, AI is here to stay and keeping up to date on its capabilities can save you a lot of trouble in your digital future. Luckily you are already taking those steps by reading Media Minds!