LinkedIn is currently testing a new AI feature called Crosscheck. This should enable users to compare several popular AI models directly on the platform. To do this, they enter a corresponding query and receive two different answers, which are generated by competing AI models – including those from OpenAI, Anthropic and Google.
Blind test for AI models
After the user has decided which answer is better, the model behind each answer is revealed. According to product boss Hari Srinivasan, this is a kind of blind test for AI models, as Engadget reports.
Crosscheck currently only works with text, but allows an unlimited number of requests. At the same time, LinkedIn shares anonymized user data with AI companies to gain insights into the models’ performance in different professional groups.
