If you believe the critics, Google faces an existential crisis from the rise of artificial intelligence.
In one doomsday scenario, rival AI tools that can answer user queries directly will siphon off the search giant’s traffic. In another, web sites filled with spammy content produced by AI will flood the company’s results with so-called slop.
But Liz Reid, Google’s vice president of search, dismissed all the pessimism, arguing at Fortune’s Brainstorm AI conference in San Francisco on Tuesday that everything is under control. The key is for Google to adopt more AI itself to make search “effortless,” opening the door to more queries.
“There are not just billions of questions that people come to Google for, but billions of questions that people don’t even ask,” Reid said on stage at the conference this week. “And they don’t ask because it’s too hard. They don’t ask because they’re not confident they can get that answer.”
Google is scrambling to add more AI tools to its search engine to counter a long list of rivals including OpenAI, Anthropic, and AI search engine Perplexity. Their tools, along with those of other more specialized AI chatbots, are gaining popularity, particularly among younger users. The less they use Google, the less the company can cash in from online ads. Some Google investors worry that it's a recipe for slowing growth.
One of Google’s responses has been to add what it calls “AI overviews” to the top of its results page for many queries. The tool provides an instant answer that’s pulled from a number of sources, without users having to click on any links. On Wednesday, the day after Reid's talk, Google unveiled the latest version of its AI model, Gemini 2.0, which will power AI search overviews and be integrated into other popular Google features.
Another major threat Google faces is the increasing number of websites filled with low-quality content produced by AI. The cost of creating such sites is so low and the practice so potentially pervasive that they risk overwhelming Google’s results and causing users to flee elsewhere for more reliable information.
Reid said Google’s fight against spammy content infiltrating its results has been going on since the beginning of search. Furthermore, she argued that AI-generated content isn’t by itself intrinsically bad. For example, people can use AI to help write amazing information, she said. The problem comes into play when people try to use AI to create huge volumes of content and that Google will try to combat it for years to come, according to Reid.
At the same time, the more Google adopts AI in search, the more web publishers are growing nervous. Google’s AI overviews, to a point, make it unnecessary for users to click on links, thereby strangling web publishers of traffic and any ad revenue that comes with it.
Reid cast AI overviews as merely a place for users to get started in their research. Additionally, many users want to get the perspectives of others, which AI overviews doesn’t really provide. In the end, however, Reid didn’t really address whether Google is sending less traffic to publishers. Instead, she gave a glass half-full answer, saying that the traffic it does funnel to publishers is higher quality and that those users spend more time on those sites.
Read more coverage from Brainstorm AI:
The debate over open versus closed AI models is ‘ridiculous,’ Meta executive says
