Python is widely used in AI and data science but lacks performance and low-level control for effective GPU utilization, leading developers to use NVIDIA’s CUDA framework, creating complexity and fragmentation. Mojo is a new language that aims to blend Python’s simplicity with C’s performance and Rust’s safety while offering a vendor-neutral GPU programming method. It’s being developed by Chris Lattner, who is known for LLVM, the Clang compiler, and Swift. Chris is the CEO and Co-Founder of Modular AI, the company developing Mojo, and he discusses his engineering journey and work on AI infrastructure and Mojo in this episode.
Kevin Ball, or KBall, is Mento’s vice president of engineering and an independent coach for engineers. He co-founded and was CTO for two companies, started the San Diego JavaScript meetup, and organizes AI discussions through Latent Space.