OpenAI is testing SearchGPT, a temporary prototype of new AI-based search features. According to the company, the tool will provide quick and timely answers from clear and relevant sources.
We’re testing SearchGPT, a temporary prototype of new AI search features that give you fast and timely answers with clear and relevant sources.
We’re launching with a small group of users for feedback and plan to integrate the experience into ChatGPT. https://t.co/dRRnxXVlGh pic.twitter.com/iQpADXmllH
— OpenAI (@OpenAI) July 25, 2024
The tool has been launched for a small group of users. There is an option to apply for participation in the tests on the website.
Successful features of SearchGPT are planned to be integrated into ChatGPT in the future.
“Searching for answers on the internet can be exhausting, often requiring multiple attempts to get the desired results. […] SearchGPT will quickly and directly answer your questions, providing up-to-date information from the internet and giving clear links to relevant sources,” states the OpenAI announcement.
During the search process, users can ask clarifying questions. The AI accumulates context after each query. The interaction resembles a dialogue with a human, the company claims.
SearchGPT cites and references information sources during searches.
“The answers have clear attribution and links so users know where the information comes from and can quickly get even more results in the sidebar with links to the source,” notes the ChatGPT developer.
OpenAI emphasizes SearchGPT’s focus on search, distancing the tool from the company’s generative AI training. Websites with the necessary information source may be displayed if they were not used for training the AI.
Earlier, media reported on OpenAI’s work on a new technology called Strawberry, which “will significantly improve the reasoning process of AI models and give them the ability to plan actions ahead.”
In July, the startup introduced a new artificial intelligence model, GPT-4o mini.
That same month, it was revealed that OpenAI is developing a five-level scale for assessing AI development, with ChatGPT placed between the first and second levels.