Nathan Grayson on Google AI: “They launched a busted product” to beat OpenAI to the punch

Nathan Grayson, a veteran gaming journalist and co-founder of Aftermath, slammed Google AI Overview in an exclusive interview. He called it “terrible” and a “busted product.”

A Rushed Job with Rough Edges

Grayson called it “mind-boggling” that Google would release the AI Overview in its current state. He pointed out several instances where the AI totally whiffed, like recommending seriously messed-up solutions to everyday problems.

Nathan Grayson: “There [was a] question about “How do I get pizza to stick or cheese to stick to the pizza I’m making?” And they’re like, “Well, you can add 0. 5 liters of glue.” No, that would probably kill someone. But Google’s telling you to do it.”

Pingal of Spiel Times: “The AI Overview says that eating ass can boost your immune system, which is great, but not really helpful I believe.”

Nathan Grayson: (laughs) Yeah, not particularly. That is very funny though.

According to Grayson, Google rushed the product out the door in a panic, fearing competitors like OpenAI would beat them. Basically, being first mattered more than quality.

“So what they launched was basically a busted product to be first, which is a horrible set of motivations. That’s not the way to make something good.”
– Nathan Grayson, Co-founder of Aftermath

Google’s Gamble

Despite the AI Overview’s flaws, Grayson acknowledged Google’s advantage—they’re everywhere. He suspects Google is betting on its massive reach to ensure the AI’s success, even if it means sacrificing quality.

“But because they’re Google, they’re like, “Well, if we put it out there, then at least like there’s something out there and, you know, Google is ubiquitous. So our thing will be everywhere and then through being everywhere, we win.” That’s not really winning. That’s just ruining your own product because you’re scared.”
– Nathan Grayson, Co-founder of Aftermath

Here are several widely reported instances where Google’s AI provided inaccurate or even harmful responses:

• Eating rocks: The AI Overview suggested eating at least one rock a day in response to a query about dietary recommendations.
• Jumping off the Golden Gate Bridge: In response to a query about depression, a fake screenshot showed the AI Overview suggesting jumping off the Golden Gate Bridge.
• Barack Obama’s religion: The AI Overview falsely claimed that Barack Obama was Muslim.
• Cats on the moon: The AI Overview falsely stated that astronauts had met cats on the moon.

Google’s AI overview is giving false, misleading and dangerous answers

The age of information is over, good luck to you all

awakenedspecies pic.twitter.com/q6lyWJyWb2

— Jack Straw (@JackStr42679640) June 1, 2024

For Nathan Grayson’s complete insights on journalism and the gaming industry, read the full interview here.

If you’d like to support our small independent company, please consider making a donation here. Your contribution will help us stay afloat and continue our work.