Once upon a time, search engines like AltaVista, Lycos, and Hurray controlled the internet. At that point, Google changed all that with its better search innovation. Since around 2002, Google has been the go-to search engine, and it’s become all the more notable. StatCounter indicates it presently holds a massive 91.6% worldwide search market. Despite many predictions, no other search engine has beaten Google.
About ten years ago, Google’s CEO and Executive Chairman Eric Schmidt said that a competitor to Google would eventually come along. He acknowledged that innovation, like a garage startup, often comes from unexpected places. The key is that the next big thing won’t copy what Google does; it will bring something entirely new, just as Google revolutionized search in its time.
Could OpenAI’s ChatGPT be the big change we didn’t see coming?
Why it is important: OpenAI’s CEO, Sam Altman, figures it would be cool to mix large language models (LLMs) with search innovation, possibly reforming how we search for and ponder data. While Google still holds a strong position in search, it’s also exploring a similar concept with its Search Generative Experience.
In a recent interview with Lex Fridman, Altman shared his thoughts on Google, search, LLMs, and more.
Altman kicked off by calling the current Google search experience “boring.” He’s not interested in copying Google but rather wants to reinvent how we find information, a process that has stayed largely unchanged for over two decades:
“…if we can build a better search engine than Google, then sure, that’s great. But I think there’s a bigger opportunity here. Google gives you 10 blue links and some ads. That’s one way to find information, but what excites me is finding a better way to help people find, act on, and synthesize information. ChatGPT is a step in that direction for certain tasks, and we hope to expand its capabilities further.”
He emphasized that creating a better version of Google’s search engine isn’t interesting. The goal is to help users get the information they need efficiently and creatively, whether through synthesis, creation, or guidance. Many have tried to outdo Google in search, but it’s a complex technical, branding, and ecosystem challenge. Altman believes the world doesn’t need another Google clone.
It’s a sentiment that Eric Schmidt’s prior statement – genuine development won’t come from imitating Google but from reevaluating how we approach search.
Duane Forrester, VP of Industry Insights at Yext, believes there’s a major change occurring in the manner we do online searches:
“Why bother competing with Google’s search? Why not provide an ad-free search since you’re already paying for ChatGPT? It’s easy to shake up the whole system – change how it works.”
LLMs + Search.
What’s exciting is that Altman thinks it would be awesome to combine ChatGPT with Search.
In his interview, Altman mentioned:
“We’re curious about how to do that effectively. That would be a neat thing to achieve.” “I don’t think anyone has figured it out yet. I’d love to make it happen. I think that would be fantastic.” There have been rumors about ChatGPT working on a web search tool. I’ve been skeptical about ChatGPT competing directly with Google in the usual search game, but what Altman is discussing here isn’t a new version of Google. It’s something different.
OpenAI isn’t trying to copy Google. But Altman believes they’re not quite ready to merge LLMs with Search to a high enough standard, although he’s definitely interested in reaching that point.
On a side note, ChatGPT reached a new traffic peak in the U.S. with 1.6 billion visits in February, as reported by SimilarWeb.
Altman dislikes ads.
Back in the day, Google was famous for having very few ads. But Altman thinks that’s changed:
“I’m not a fan of ads, just from a design standpoint. Ads were necessary to kickstart the internet, but it’s a temporary solution. Nowadays, we have better options. I prefer that people pay for ChatGPT to get unbiased answers, free from ads. Sure, there might be a way to include ads in LLMs without bias, but it’s easy to imagine a future where ChatGPT pushes products or ideas for profit.”
He appreciates OpenAI’s straightforward business model where users pay directly:
“Our business model is simple, and I like that. I know I’m not the product; I’m a paying customer. I’m not a big fan when I use ad-supported platforms like Twitter or Facebook. I think it’s getting worse, not better, with AI in the mix.”
Altman believes OpenAI can fund its needs without resorting to ads:
“We have a solid business that can cover our costs without ads. There should be better advertising methods that don’t mess with content or manipulate truth for advertisers.”
But there are challenges. Brett Tabke, CEO of Pubcon, pointed out Google’s advantage in data:
“Google has a wealth of data that OpenAI doesn’t. They’ve digitized millions of books with vast StreetView imagery, YouTube content, Android data, Gmail insights, and Chrome browsing data. This data advantage is a significant hurdle for OpenAI to overcome.”