Jack Nystrom

About

Check out my various projects, writing, and my resume/experience!

---

A bunch of things in no particular order:

• I spend most of my time in the basements of Baskin Engineering, working on an EV racecar
• Passionate about high quality software engineering and electronics, will talk for hours (be a nerd about something, don't be boring!)
• Mountain biker, but mostly as a social activity (I'm still pretty mediocre)
• You won't find me coding in anything other than Neovim
• Big fan of compiled programming languages that still expose the details without compromising abstractions and safety. Hopeful that Zig's new IO interface and the colorless-async dream will have useful event-loop applications in embedded programming
• I use NixOS not because it saves me time in the long run but because I absolutely love getting to control and understand the pieces of my OS and the reproducibility of everything down to specific program build flags and kernel parameters. And because having isolated tooling is awesome
• I believe in cities designed around pedestrians and bikes over 50mph expressways and 8ft tall lifted trucks. E-biking is my favorite part of the day, give it a try! People would be healthier and happier.
• I am strongly against AI agents. LLMs are a super useful tool, but the way Big Tech is obsessed with AI and with replacing human labor is nothing short of awful. Largely, AI is in direct opposition with everything we should be fighting for. Allow me to explain.
  • The code that AI produces is mostly unuseable and/or unmaintainable without skilled human review and manual cleanup. In the hands of the right person this can produce a viable output, but in most real world cases the output of this AI-centric workflow is a product with strictly worse quality and poorer maintainability than the same product if it had been made purely by humans. In a world where software quality is essential to the everyday lives and safety of billions of people, and where software is already so frustratingly low-quality in so many cases, I am very concerned by the direction we're headed. Of course, CEOs and investors love anything that promises a faster product to market. But the world needs us to focus on quality, not getting out the door faster.
  • Using an AI agent is outsourcing your intelligence to a tool owned by one of ~3 billionares. We sacrifice our agency and critical thinking in the name of productivity, and hand over the intellectual ownership of our product to a tool run by our greatest enemies. We introduce a link in the chain of information flow with the ability to wholely manipulate it in whichever way they want, invisibly to the user. Furthermore, the sacrifice of our learning to artificial intelligence increases our dependence on it, attaching a subscription fee to the very basis of producing work. Who does this "productivity" really benefit? Hint: it's not us. (Meanwhile, Nvidia continues to gladly sell the shovels...)
  • I believe that the best way to make use of AI is as a glorified conversational search engine to help you learn about well-known topics. It has ingested basically all of the publicly available human knowledge at this point, and is extremely good at articulating it for you. What you should NOT use it for is: designing APIs and interfaces, deriving new solutions to unsolved problems, designing your product, and especially not autonomously coding anything.
• I am very concerned about the state of US politics, our dysfunctional government, the scale of corruption, the political polarization, the inability to pursue any remotely radical progressive change, the intentionally sensationalistic media, the voter suppression, the ever-increasing national debt and defense budget, the administration full of guilty but unstoppable narcissists, and the embarassment of a national leader
• Induction stoves are awesome
• Huge fan of Jazz and harmony and think that things like this are peak music (give it a chance)