Takato Honda, a Japanese data scientist, has recently launched a new open-source programming language called "Sui" (Sui), aimed at revolutionizing the accuracy of code written by large language models (LLMs) and claiming that it can enable LLMs to generate code with 100% accuracy.
According to the information, the name of the language comes from the Japanese aesthetic concept of "Sui," meaning **"refined" and "eliminating redundancy."** Its design philosophy is to retain only the essentials. The core design principles of Sui include: ensuring zero syntax error rate in structure; achieving zero spelling errors by using numbers as variables instead of names; guaranteeing line-level independence, meaning each line of code is completely independent; and being designed as a pure logic language for computation, allowing the user interface to be freely combined with any existing framework (such as React, Vue, etc.) for implementation.
Takato Honda believes that once LLMs learn Sui, its token efficiency optimization will make it more efficient than existing languages, bringing breakthroughs to the field of AI-assisted programming.
Link: https://github.com/TakatoHonda/sui-lang
