OpenAI uses Reddit content to train artificial intelligence (AI) models.
This is one element of a new partnership formed by the AI company and the social media platform, Reddit said in a blog post on Thursday (May 16).
According to the post, the partnership will allow OpenAI to access Reddit's Data API to provide content from Reddit to ChatGPT and new products to serve structured and unique content from the social media platform in real time.
“This will allow OpenAI's AI tools to better understand and present Reddit content, especially recent topics,” the post said.
As another part of the partnership, Reddit will build on OpenAI's AI model platform and provide new AI-powered features to Reddit users and moderators, the post said.
Additionally, OpenAI plans to advertise on Reddit.
“Reddit is the Internet's largest chronicle of authentic, relevant, and always up-to-date human conversation about everything,” Steve Huffman, co-founder and CEO of Reddit, said in a post. It has become an open archive.” “Including this in ChatGPT maintains our belief in a connected internet, makes it even easier for people to find what they're looking for, and makes it easier for new viewers to find our community on Reddit.”
OpenAI Chief Operating Officer Brad Lightcap added in the post: “We are excited to partner with Reddit to enhance ChatGPT with unique, timely and relevant information and explore the potential to enrich the Reddit experience with AI-powered features. ”
The announcement follows an agreement between Google and Reddit that allows Google to use Reddit's content to train its AI models.
In March, it was reported that Reddit received a letter from the Federal Trade Commission (FTC) regarding data licensing practices related to training AI systems.
The letter tells Reddit that FTC staff is conducting a private investigation focused on the company's sale, licensing, and sharing of user-generated content with third parties for the purpose of training AI models. Ta.
Reddit said in a filing at the time that its data licensing efforts were still in the “early stages” and that such efforts are in line with its values and respect the rights of its users.