Lingguang 2.0 has been online for two weeks, and the cumulative number of managed transactions has exceeded 3 million.
On November 11, Ant Group’s personal AI assistant Lingguang announced the latest data: Since the launch of Lingguang 2.0 on October 27, the cumulative number of managed transactions has exceeded 3 million, and the cumulative number of tasks has exceeded 11 million. This is another disclosure of transaction custody data since the launch of Lingguang 2.0. On November 5, Lingguang announced that it had hosted more than 1.16 million transactions within a week of its launch, performed more than 3.6 million tasks cumulatively, and had more than 20 million users. Different from traditional AI product indicators such as user size and number of conversations, transaction hosting volume measures how many things a user has handed over to AI for continuous follow-up. Its growth reflects that the use of personal AI assistants is extending from instant question and answer and single task execution to long-term transaction management. The key to this change is that AI shifts from responding to instructions to continuously fulfilling contracts. Traditional AI assistants usually complete an interaction after the user initiates a question or instruction. The "Help Me Manage" function launched by Lingguang 2.0 allows users to hand over matters that require long-term attention to AI after authorization, which will continuously monitor the status and proactively remind and perform corresponding operations when conditions are met or critical nodes are reached. For example, parents can ask Lingguang to continue to follow the DingTalk class group and summarize important notifications and to-do items; consumers can host ticket price change monitoring and automatic membership renewal reminders; users can also connect smart home devices to let AI automatically perform operations based on preset conditions. What these scenarios have in common is that the task does not end in a conversation, but requires the AI to maintain attention and take timely action without continued user intervention. Supporting this capability are long-term task scheduling, cross-session context management and external service connection capabilities. As of November 3, Lingguang has connected to more than 40 external connectors such as Alipay, DingTalk, Quark, Feishu, etc., expanding the scope of AI services to work,
Read more







