While ChatGPT caused a lot of noise when it was announced in 2022, the arrival of ChatGPT Search less than a year ago has Google worried... While the ChatGPT model focused on a set of documents, ChatGPT Search (or SearchGPT) goes much further by generating sourced answers in real time.
While ChatGPT already represented an initial threat to Google's monopoly in search, this new feature d’OpenAI has been developed perceived as a real declaration of war for the American giant.
41.3 ChatGPT Search users in Europe
According to the latest figures released by Statista on search engines, Google is still far ahead of its competitors like Bing, but this share could soon decrease...
In order to comply with the DSA (Digital Services Act) in Europe, OpenAI must regularly publish information about the use of its services on the continent. Thanks to these reports imposed on the largest global platforms, it is now possible to more precisely monitor their usage statistics.
In a latest report communicated by OpenAI Ireland Limited this Monday, April 21, the figures communicated by the company are showing very strong growth. If we could already count 30.1 million users, there are already A few months later, the official figure of 41.3 million monthly users over the last six months remains even more impressive.
8% of users choose ChatGPT according to a survey
Faced with the growth of ChatGPT, traditional players like Google will have to redouble their efforts to remain competitive. In a survey published by Fortune in 2024, 8% of those surveyed said they have more confidence in ChatGPT than in Google.
While these figures show that OpenAI is still a long way from reaching Google, the gradual rise in usage trends is worrying the search engine. Indeed, while Google has long been criticized for lagging behind in artificial intelligence, the latest announcements regarding Google Gemini seem encouraging.
Furthermore, another study conducted by researchers could make the path more difficult than expected for OpenAI...
67% errors on ChatGPT Search
In order to better understand user usage, researchers recently conducted a study on the results provided by ChatGPT Search published by Ars Technica. With a 67% error rate on the articles mentioned by OpenAI, this shows that the reliability of the tool is not yet there.
A second study, this time conducted by CJR, also showed that ChatGPT Search's news-related results were not reliable enough. While future updates to ChatGPT Search might address this shortcoming, will it be enough to dethrone Google?
0 Comments