The company then went on to strike provides with most important tech firms, along with a $60 million settlement with Google in February 2024 and a partnership with OpenAI in May 2024 that in-built Reddit content material materials supplies provides into ChatGPT.
Nonetheless Reddit prospects haven’t been totally happy with the provides. In October 2024, London-based Redditors began posting false restaurant choices to handle search outcomes and defend vacationers away from their favorite spots. This coordinated effort to feed incorrect information into AI strategies demonstrated how explicit individual communities may intentionally “poison” AI instructing info over time.
The potential for trouble
Whereas it’s tempting to lean intently into generative AI know-how whereas it is presently trendy, the change might also characterize an issue for the company. For example, Reddit’s AI-powered summaries may greater than doubtless draw from inaccurate information featured on the place and provide incorrect selections, or it’d draw inaccurate conclusions from appropriate information.
We’ll regulate Reddit’s new AI-powered search instrument to see if it resists the sort of confabulation that we have now seen with Google’s AI Overview, an AI summary bot that has been a important failure to this point.
Advance Publications, which owns Ars Technica guardian Condé Nast, is a crucial shareholder of Reddit.