On May 24th, the Unisound Shanhai Model and Achievement Conference was held in Beijing, where the ten core capabilities of the Shanhai Model were demonstrated, and a series of product applications for different industries were announced.
Huang Wei, the founder and CEO of Unisound, stated that the release of the Shanhai Model is an important milestone in the upgrade of Unisound’s AGI technology architecture. Based on the Shanhai Model, Unisound will develop an AI 2.0 solution in the form of the MaaS (Model as a Service) mode. Building upon its general capabilities, Unisound will enhance its abilities in industries such as IoT and healthcare, providing customers with smarter and more flexible solutions, and opening up greater commercial opportunities for AI technology. Additionally, Unisound will continue to upgrade the capabilities of the Shanhai Model, with the goal of matching the general capabilities of ChatGPT within this year, and surpassing GPT4 in multiple vertical domains such as healthcare, IoT, and education.
During the conference, the ten core capabilities of the Shanhai Model were demonstrated, including language generation, language understanding, knowledge question answering, logical reasoning, coding ability, mathematical ability, and security compliance. In addition, there were three industry-specific capabilities: plugin extensions, domain enhancement, and enterprise customization. Unisound explained and demonstrated these capabilities through speech input and real-time interaction.
As one of the most representative abilities of generative AI, language generation is one of the foundational capabilities of the Shanhai Model. It can not only generate various fluent and coherent texts, including news, essays, novels, emails, poetry, and couplets, but also support various language generation tasks in multiple languages, styles, and genres, including creative writing, summarization, and translation. Furthermore, it supports controlled text generation under various constraints.
In terms of language understanding, the Shanhai Model can comprehend complex ideas by combining context, common knowledge, and specific knowledge, enabling a deep understanding of the true meaning and underlying emotions of statements. In knowledge question answering, the Shanhai Model demonstrates impressive knowledge depth, breadth, and interdisciplinary integration. The three capabilities of logical reasoning, mathematics, and coding highlight the thinking chain abilities behind the Shanhai Model. Currently, the Shanhai Model is capable of observing, comparing, analyzing, synthesizing, abstracting, summarizing, judging, and reasoning about things, expressing its thought process accurately and systematically. Its mathematical and coding abilities are also continuously evolving.
Moreover, the Shanhai Model ensures the legal compliance of its outputs and can provide positive guidance, avoiding potential security risks.
While large models serve as the foundation of knowledge and capabilities, the plugin extensions, domain enhancements, and enterprise customization capabilities of the Shanhai Model further expand the boundaries of its abilities, meeting the higher demands of various industries for flexibility, generality, and practicality.
In the grand wave of AGI, the deep integration of large models with specific application scenarios has become inevitable. To improve the level of application implementation in specific scenarios, Unisound adheres to its consistent U+X strategy, which combines U (AI technology and product capabilities) with deep integration of X (industry application scenarios) to address the deep-seated issues in various industries.
In the medical field, leveraging years of deep cultivation in the healthcare industry and based on accumulated past data and experience, Unisound has comprehensively upgraded the intelligence level of various products in its medical business line by relying on the Shanhai Model. They have released three major medical product applications: surgical medical record writing assistant, outpatient medical record generation system, and intelligent commercial insurance claims system. This advancement enables a leap from being an assistant to becoming an expert.
In the sales scenario, Unisound combines the Shanhai Model to upgrade its Yunbei sales management system. Through a four-step process consisting of customer profiling, intelligent review of sales calls, high-quality sales dashboards, and generated follow-up tasks, Unisound keenly identifies the strengths and areas for improvement in the sales process. This enables a more accurate understanding of customer needs and facilitates efficient customer acquisition.
In the knowledge management scenario, traditional internal knowledge management processes in enterprises often encounter issues such as numerous answer versions, complex and abundant information, and imprecise retrieval when employees search for solutions to problems. Addressing these pain points, Unisound upgrades its existing knowledge management service system by relying on the Shanhai Model and creates an enterprise-level solution called “New Bing.” It provides concise answers to assist employees in understanding internal technical documentation and offers precise traceability for each response.
Looking at the smart IoT scenario, Unisound plans to comprehensively upgrade its core smart IoT products in the future, deeply integrating them with the Shanhai Model. This integration aims to create a true personal assistant, transforming it from mere command-based interactions to human-like conversations and truly linking with the IoT ecosystem and services.
On the day of the conference, Unisound entered into strategic partnerships with CSCEC, JD Technology, and 360. They initiated in-depth collaborations with these partners to promote the practical application of the Shanhai Model in various fields, collectively stepping into the wave of the AGI era.