Coursera Spring Sale
40% Off Coursera Plus Annual!
Grab it
Explore a 16-minute conference presentation introducing Meaning-Typed Programming (MTP), a revolutionary paradigm that simplifies AI-integrated software development by abstracting large language model (LLM) integration through intuitive language-level constructs. Discover how MTP addresses the complexity of manually crafting prompts and processing outputs in AI applications by leveraging the semantic richness of code to automate prompt generation and response handling. Learn about the three core components: the "by" operator for seamless LLM invocation, MT-IR (a meaning-based intermediate representation for semantic extraction), and MT-Runtime (an automated system for managing LLM interactions). Examine the implementation of MTP in Jac, a programming language that supersets Python, and review performance results showing that developers completed tasks 3.2× faster with 45% fewer lines of code compared to existing frameworks. Understand how MTP demonstrates resilience even when up to 50% of naming conventions are degraded, proving its robustness to suboptimal code conditions. Gain insights into this research presented at OOPSLA 2025 by a team from the University of Michigan and Jaseci Labs, which is available as part of the open-source Jaseci project under the byLLM module, complete with artifacts that have been evaluated as reusable with reproduced results.
