We watch AI reshape investing at Hum every day, and one thing keeps proving true: the deterministic systems the industry spent decades building aren't legacy, they're the foundation AI runs on. Blair Silverberg breaks it down through a deceptively simple question: what actually happens when a calculator versus an LLM adds 2+2? A sharp read for anyone thinking about where durable value lives in the AI stack.
Ryan Shmeizer and Jake Saper's great work has inspired me to share some of the thoughts I am having as I watch the Hum Capital team use AI to invest everyday. The first is that rules-based, deterministic code can be highly valuable even in a world of AI. Perhaps nothing illustrates this better than thinking through the physics of doing 2+2 on a TI-83 vs. your favorite LLM! All of this is mute, however if Naveen Rao finds a way to give us free computation :)