Skip to content
aiqbench takeAI Products

aiqbench analysis

Ant Group's Lingguang 2.0 beta marks a strategic pivot from a reactive chatbot to an AI-native personal agent featuring long-term memory and proactive task execution. This evolution addresses critical challenges in multi-step reasoning and real-world utility, positioning the model as a direct competitor to Meta's Muse within the Chinese market.
Source covering36Kr
What this isA model's reading of the article and related coverage, written for aiqbench. Not the publisher's words.

Ant Group's 'Lingguang' Begins Small-Scale Beta, New Version Targets 'Meta Muse'

36Kr1 min read

Ant Group's AI product 'Lingguang' has recently launched a new version for a small-scale beta test, with the update targeting Meta's 'Muse'. Public information indicates that in September this year, Meta introduced its personal agent, Muse, which can intervene in users' daily lives and continuously follow up on tasks assigned by users. Sources close to the project revealed that preparations for Lingguang began about six months ago, during which time multiple versions were iterated upon, with plans to fully open access this month. The source also noted that the core of this upgrade marks a shift for Lingguang from an AI assistant capable of chatting and generating applications to an AI-native personal agent equipped with long-term memory and the ability to proactively advance tasks. The new version of Lingguang can also execute functions such as sending messages via DingTalk or controlling smart home devices by invoking Skills and MCP. It is reported that the beta test cases for Lingguang 2.0 focus on scenarios involving long durations, numerous variables, and situations difficult for ordinary people to monitor, such as tracking market fluctuations for office workers, helping parents manage DingTalk clock-ins, or rerouting travel plans in response to flight delays or crowd control at scenic spots. (Sina Technology)

Read the full article at 36Kr

Related coverage