OpenAI has announced a prototype version of SearchGPT, an AI-powered search engine designed to compete with major players like Google Search and other existing search engines.
This development marks a significant shift in how users may interact with search engines, leveraging advanced AI technology for more efficient and relevant search experiences.
SearchGPT offers two key advantages over traditional search methods and even OpenAI’s ChatGPT:
The visual elements of the search results, including the citation of sources, add to the clarity and reliability of the information provided.
OpenAI has not specified whether SearchGPT will integrate with existing search engines like Microsoft Bing. However, the company has highlighted the potential for real-time information retrieval to streamline the search process.
This feature reduces the effort required to find relevant results, a common challenge with traditional search engines.
OpenAI’s statement emphasizes the goal of enhancing their models’ conversational capabilities with up-to-date web information, making the search process faster and easier for users.
The launch of SearchGPT comes as other AI companies, including Google and Perplexity, are also enhancing their search capabilities.
Google has recently introduced expanded AI Overviews and AI-organized search results, utilizing Gemini’s multi-step reasoning. Perplexity has integrated conversational search and the Wolfram|Alpha engine for academic research.
OpenAI’s initiative to introduce SearchGPT directly responds to these advancements, aiming to offer a competitive alternative in the AI search engine space.
Access to SearchGPT is currently limited, and users must join a waitlist for early access. However, OpenAI plans to incorporate the best features of SearchGPT into ChatGPT in the future, indicating a broader vision for integrating advanced search capabilities across their platforms.
OpenAI has also clarified that SearchGPT is designed to support publishers by prominently citing and linking to them in search results. This approach aims to provide transparent attribution and foster better connections between users and content creators.
The use of data from publishers and content creators to train AI models has been a contentious issue. OpenAI has addressed these concerns by separating the data used for SearchGPT’s search results from the data used for training its generative AI models.
The OAI-SearchBot crawler will be used to surface relevant content in search results, independent of the training datasets.