Have you ever wondered how ChatGPT answers your real-time questions? While rivals like Gemini and Grok have their own internet platforms to scrape data, OpenAI’s ChatGPT doesn’t have such luck. Based on a new report now, it is said that Sam Altman’s ChatGPT might be relying heavily on Google to bring up answers. This is ironic considering how OpenAI’s ChatGPT is dead set on replacing Google as the go-to search engine.
The report, which comes from The Information, reveals that ChatGPT is relying on a custom tool that feeds data from Google Search. This comes across as fascinating considering how Google allegedly denied OpenAI’s request for direct access to its search index, which is a treasure trove of web information for any AI chatbot.
ChatGPT feeds off Google Search data
According to the report, OpenAI is using a paid web-scraping service called SerpApi to extract up-to-the-minute search results from Google. This method allows ChatGPT to deliver fresh, real-time information on topics like news, financial markets, and sports – all the topics where its own AI models have not yet been fully developed. The findings were supported by a former Google engineer, Abhishek Iyer, who demonstrated that ChatGPT could access information from dummy pages that were exclusively indexed by Google.
While OpenAI CEO Sam Altman has publicly stated he no longer uses Google Search for his search purposes, his company’s reliance on Google’s data for its own products suggests a dependency. During a Google anti-trust trial, Nick Turley, the head of ChatGPT, confirmed that OpenAI had sought access to Google’s search index but was turned down. Turley also noted that search results from Microsoft’s Bing, another data source for ChatGPT, had “significant quality issues,” further emphasising the value of Google’s data.
OpenAI is yet to issue a public statement regarding the allegation.
Can ChatGPT thrive without Google?
Without an up-to-date database, any AI chatbot would not have information about what’s happening in the world. Previously, ChatGPT’s older versions used to be on the back foot, considering the dated database they used for training. Google’s Gemini and xAI’s Grok, on the other hand, have always relied on Google Search and X, respectively, as their data source, thus giving them a competitive advantage.
