OpenAI revealed SearchGPT, an innovative search engine powered by artificial intelligence. This sophisticated tool, currently in phase di test with a small group of users and selected publishers, promises to offer an unprecedented browsing experience, providing results not only in the form of links, but enriched by arguments generated by an advanced linguistic model.
SearchGPT: what the OpenAI search engine is and how it works
SearchGPT's announcement comes as a surprise, considering that rumors about such a project had circulated in May, but the company had then focused attention on the "Springs Update" event, which mainly focused on the model GPT-4 and new vocal capabilities. However, those interested in testing this technology can sign up for one waiting list, waiting to be selected for early access.
The company is convinced that by integrating the capacity conversational of your models with real-time data coming from the web, it will be possible to make the search experience faster and more efficient, as highlighted by the famous writer Jorge Luis Borges in his story "The Library of Babel", in which he imagines an infinite library containing all possible books, symbol of infinite human knowledge.
SearchGPT stands out for its ability to provide updated answers, drawing on current information from the web, but at the same time maintaining transparency by clearly citing and linking the sources used. This feature reflects OpenAI's commitment to promoting a balanced digital ecosystem, where publishers have a greater control over the presence of your content within the search engine.
It is important to note that the SearchBot, used to navigate and link to sites in search results, is not a traditional crawler. It does not extract data from websites to train OpenAI generative models, but it is limited to index and present relevant content. This distinction is essential to reassure publishers that their rights are protected and content is correctly attributed.
The advent of SearchGPT, however, raises questions about the delicate balance between technological innovation and protection of artists' rights, photographers and writers, who have raised concerns in recent years about using copyrighted material to train generative AI. This debate also extends to publishers, who fear a potential decrease in traffic to their sites due to AI-generated all-encompassing search results.
Now the challenge for OpenAI and for the entire generative AI sector will be to find a compromise that favors technological progress without compromising the existing digital ecosystem. The company is committed to using the SearchGPT prototype to delve deeper into these issues and share the results with the community, in hopes of finding solutions that meet the needs of all parties involved.
