Rejecting AI's nonsense! Microsoft Corporation (MSFT.US) launches the Critique system: dual-model collaboration achieves self-correction.

Microsoft Corporation (MSFT.US) unveiled a series of new artificial intelligence upgrades on Monday, launching a new multi-model deep research system called "Critique." According to the company's CEO Satya Nadella, the core of this system adopts an innovative dual-model collaborative architecture, breaking down the traditional single generation task into two separate stages: "generation" and "evaluation." In this system, the first model is responsible for initial task planning, information retrieval, and drafting, while the second model acts as a senior reviewer, specifically verifying the accuracy of facts, checking the logical chain, and refining the final report. This self-correcting mechanism achieved through multi-model interaction aims to fundamentally overcome the common "illusion" phenomenon in existing AI models and significantly improve the depth and quality of research reports. Notably, the Critique system demonstrates high levels of openness and compatibility, no longer limited to a single vendor's single model. According to official technical details, the system integrates OpenAI's GPT series, Anthropic's Claude series, and Microsoft Corporation's proprietary Phi series models, enabling cross-vendor intelligent collaboration. For example, the system can utilize the powerful creative generation ability of the GPT model to produce an initial draft, which is then subjected to rigorous logic auditing by the Claude model. Based on the latest benchmark test data released by Microsoft Corporation, Critique has outperformed similar single-architecture products on the market in core dimensions such as breadth, depth, and quality of expression, demonstrating strong competitive advantages. In conjunction with the release of Critique, Microsoft Corporation also introduced the complementary functionality system "Council," further enhancing the application scenarios of multi-model collaboration. The Council system allows users to run multiple AI models from different sources simultaneously for the same complex instruction on the same interface. By introducing an independent third model as a "neutral arbitrator," the system can automatically identify and summarize the similarities and differences in the results produced by each model, helping professional researchers capture unique insights that a single model may overlook, thus greatly enhancing comprehensive decision support. "To use this feature, select the Model Council from the Researcher's model selector. Council will simultaneously run models from Anthropic and OpenAI, generating a complete independent report for each model, highlighting facts, citations, and analysis frameworks that the other model may have missed or weighted differently," explained Microsoft Corporation. "After the two reports are generated, a dedicated judgment model will evaluate the reports, extract key findings, and highlight meaningful consensus or differences between the modelsincluding differences in scale, framework, or interpretationwhile pointing out each model's unique contributions." These cutting-edge technologies have already been integrated into the "Researcher" toolkit within Microsoft 365 Copilot. According to Microsoft Corporation's market promotion plan, Critique and Council functionalities are now in the early testing phase, with initial access limited to enterprise customers who have joined the Microsoft Corporation "Frontier Program." Analysts point out that with the implementation of this deep research system, Microsoft Corporation's moat in the enterprise productivity tools market will be further strengthened, signaling a shift in AI competition from a simple model parameter race to a new stage of complex system integration and logic verification.