Rewrite AI with NVIDIA

As a member of Rewriting the Code, I am working to learn more about the tech field every day. One of their “Innovator Series,” a monthly webinar for education, was hosted 2/26/2026, called called “5 Ways to Show You’re Ready to be an AI-Native Teammate” or “Rewrite AI with NVIDIA.”

Rewriting the Code Event

Notes

Krapu’s Career

  • 2009-2013 College, 2013-2014 First Job, 2014-2020 Grad School, 2020-2022 Scientist, Currently AI Eng
  • Core skillset never goes away
  • “you have to have good fundamental core skillsets, because then you’ll be prepared for new technologies”

How the world is changing

  • Code is cheap
  • Judgement is bottleneck
  • Every part of SDI will change

1. Bring Opinions to your new job
If they’re releated to code.

  • “The most valuable thing you can do isn’t being a blank slate…is it you have strong stylistic opinions about how code should be strucutred…that can be really valuable”
  • The diverse opinions of having employees is a strength
  • Open source, being experienced in a range of other things
  • “What I love to see from folks that are starting out is pushback”

2. Keep a public AGENTS.md

  • Keep a style sheet to have a ground
  • This reminds me of supreme court precedents from government class
  • “What do you think about the most, what irritates you the most, when you go into software”
  • Avoid doing this, do this instead
  • Works as a miniguide as well

3. Try more tools than your coworkers

  • “We get blind spots and forget what’s out there”
  • “Junior staff love to try new things, they don’t have these biases”
  • Perplexity and research -“ All the tools that are out there, even if you think they aren’t super germane(?), it doesn’t hurt to have those discussions”
  • People’s preferences can be a weakness

4. Own your technical career development

  • Strong, important point!!
  • Built off technical proficiency as you progress in your career, but collaborative learning happens behind the scenes
  • Have a deep interest in the learning in regards to core functionality
  • If he had to start from scratch, he’d ask peers how to prevent getting rusty/losing touch in relation to technical skills
  • if you’re serious in the long run, have this discussion with team/manager/peers

5. Be Cynical.
about your code’s ecosystem.

  • be aware of the new realities of the software ecoystem, especially risks
  • a lot of staff aren’t going to be aware as new gen is raised in a different environment
  • similar to parents picking up the phone for scammers
  • always take the responses from LMs with a grain of salt

Questions

Some people have raised concerns that AI tooling is preventing students from developing the core critical thinking skills. How should students use these AI tools in such a way that they enhance their critical thinking skills?
Answer: “When we learn, we often see a limited slice of the material,” learning is deepened by looking at more core theories, “it is very very hard for textbook publishers and teachers to understand how to structure material in the amount of time given,” and AI provides that by allowing you to tailor your experience and learning.

What do you think is important for intern/new grad to showcase during the interview?
Answer: Agency is the top of the pile, “side projects is my number one reccomendation.”

How to keep up with AI?
Answer: “Some combination on reading, seeing popular discourse, try to dip your toes into the research and see: ‘what is the hardest part?’”

How do I avoid getting lost when coding with AI? It generates so much code at once that I sometimes lose track, struggle to review it all, and don’t even know what I should study to understand the logic behind it.
Answer:

  • “I try to condition myself to be extremely picky…I’m going to be so critical and so nitpicky, and if it isn’t the style I want, I will rewrite it myself.”
  • “The amount of effort you do for each file is going to vary.”
  • “Do something stylistically with the code”

How do developers ensure they are not replaced by AI?
Answer: The #1 thing is being an effective organizor of information. Be the least biased, and that can strengthen the use of AI.(?)

You mentioned people focus more on code execution nowadays than actually learning how to design good code. I feel like I struggle with this at work. What do you do to not loose that ‘touch’ as a software engineer?
Answer: Both AI and humans thrive on explicit assumptions. Keep an appropriate amount of reading as a part of your day.

What should I study to better understand AI-generated code? School doesn’t cover enough and I feel like I need to actually know the tools and frameworks myself.
Answer: Advocate for your university programs to expose you, with Cursor or Claude Code. The textbooks aren’t written yet, and that’s the phase we are in.

Final Thoughts

Thank you to Sidney Lopez, the Sr. Program Manager at Rewriting the Code and Speaker Christopher Krapu, Solutions Architect at NVIDIA for your time and advice at this event! This was very informative, and thank you for reading. :)


All blog posts are written by hand, no AI usage.