Written by 
The AI startup company Prosocial Intelligence has confirmed its acquisition by Ant Group. The company will operate independently, but the investors have already exited.
There are reports that Ant Group has recently established a reinforcement learning laboratory, and Wu Yi, the founder of Prosocial Intelligence, has joined as the chief scientist of the laboratory.
Wu Yi responded to his joining Ant Group on social media by saying: “Recently invited by Ant to conduct research in reinforcement learning for their large models. I am continuing my teaching at Tsinghua University’s Institute for Interdisciplinary Information Sciences and working with Prosocial Intelligence. Ant has great applications for large models, and I hope we can make impactful contributions together. Thank you all for your concern, this is my unified response.”
According to informed sources cited in reports, Wu Yi had doubts about the commercial prospects of Prosocial Intelligence before and had been looking for acquisition opportunities. Previously, Ant Group had been seeking investment targets in the AI field, and both parties reached an agreement in this direction. Similar to previous acquisitions of Oladance and AIWaves Inc., Prosocial Intelligence will continue operating independently after being acquired. The sources mentioned that current investors have exited.
Informed sources revealed that after angel investing rounds, Prosocial Intelligence’s valuation had reached $80 million USD, but the acquisition price by Ant was lower than the company’s valuation. Investors such as ZhenFund and Shunwei who participated in angel investing did incur some losses; however, these losses did not exceed 50% of their initial investments.
Public information shows that Prosocial Intelligence was founded in March 2023 with a focus on providing better intelligent experiences using reinforcement learning based on large language models for more people. In a podcast from one round of ZhenFund funding previously mentioned by Wu Yi stated he hopes to build alignment technology between AI and humans through a universal reinforcement learning training platform so intelligence can be presented to people most comfortably possible benefiting as many people worldwide as possible.